Mexico - Rainfed Areas Development Project

Document of The World Bank .FOR OFFICIAL USE ONLY Report No. 12533-M STAFF APPRAISAL REPORT MEXICO RAINI AREAS DEVELOPMENT PROJECT JUNE 15, 1994 '2 A.. _(RlEiEI! , iFvre: X -h Agriculture Operations Division Country Department II Latin America and the Caribbean Regional Office This document has a restrted distibution and mnay be used by redpients only in the perfonuance of their offieiul duties. Its contents may no otherwise be disclosed without Wodd Bank a on.  Curemc Unit Nixcan Now Peso (NSI IJSSIZ,SSt t N$1 railliot (Jure 1994) FISCAL YvAk lanuarY I Dccmnber 1i '*EtQfrS AN4D NIASR Metrr Systcrn ABOMlWITONS AND ACRONY1S A(SAL Agricultural Seaor Adjustment Loan ATP Agricultural Teebrology Pmoject BANRUIRAL ational Rural Bank (Banco Naionai de C(edito Rural, CNA National Water Commission (Cornisitin Nacionai del Agua) CODEiTr Delegational Imnrinstitutional Cotntittee ott Technology Tfansfer (Comtui Delegaetonai Interiostufnonal de Trarsferetcia tde Teooltg )a CONASUPO n4atiotna Food Supplies Comnp, (Compalba Nacional de Subsistettcias CC' FL il LIE ',,e Pla-mirg ..rm _ .-:.c J P.ai.:fc .vo hstatai. DDg Rural Development District (Distrito de Desarrollo Rural DRD Decentralization and Regional Development Projeta FIRA Ttust Fund for Agriculture (Fideicomszos tnstituidos en Relaclfn con la Agricultural FIRCO Trust Fund for Shared Rtsk (TFdelcomiso de Riesgo Comipartidot GOM Govemnenr of Me-Aco IDSP Irrigation and Drainage Sector Projec INIFAP National institute for Forestry, Agriculture and Livestock Research (Instituto Nacional de Investigaciones Forestales s AgropecuariasI IPM Integrated pest mnanament MTR Mid-term reiea, NAFTA Nonbh Amencan Free Trade Agreement NAFIN Natonal fr-nite Bank iNational Finsncleral OFf4IN On-Frti amd Minor Imgaiiol NetWorks 1i1provenient Project PLAXAT Nautonaj PLn for Ramfd Agriculture Deselopnteit (pltn Natol par Ia Agnicultu Temporall PRoCAMPO -rFde Uhetliuti,n ind Direct Producer Income Suppon Progrm PROCEDE Land Ceilificatton Prograt iPrograa d Ceetifcaion de Derechos Ejidales, PROCATI Aguliture Trinaing. Extenson and Resauch Progrin (pgl d CopaCiaci6o, Asicia T&entra e Intigati6n1 PRODERETH Progranm for lnjtgrtd Rural Development of the Hurrid Trupies (Prograna, e Dearrollo Rural Inegmrl pAr e Trd6pIo Hiintedo! PRONASOL National Sohdarit) Vrogram (tPengra-a Narional de Soltduaridad) SARH Secrt of Agricutture nd tiYdraut1i Renoureis (Secretarji de Aglicolnra- y Recursos Hidrnliros) SECOGEF -ate Secreariat of Comptroller of the F ion (Secetrf de a Cotrvdorfia Gmel de Ia Federacidis) SEDESOL Seatat for Soeia DeveLopn (Seeara de Dertllo Social) SOE Statotmet of Ex Editr Estlo de Gastos)  FOR OFFICIAL USE ONLY - MEXICO RAN1D AREAS DEVELOPME?fl PROJECT CONTENTS page No. LOAN AND PROJECT SUR.ARY .... i L TME AGRICULTURAL SECTOR . . . 1 A. SectorujOutputandGrowth .... B. Straegy fo a Transition to a Market-oriented Agdculture . ....................... 2 C. Ihe Adjustment to NAFTA .. . 4 H. RAINFED AGRICULTURE .......................- 6 A. Bariers to Growth in the Rainfed Ara.. 6 m. THE PROJECT ............................ 8 A. Rationale for Bak involvemnt . . ............................. S B. Lessons from Bank Experience. . . . 9 C. Project Objectives .......................................... 11 D. Project Description . . ............................ . 11 1. Project Components ......... ............ .................... 12 2. Project Costs and Financing ...... . . . . . .... 21 E. Project Organizaion and Management .23 1. Coordinating Agency .23 2. Project mplementation .24 3. Procuement .25 4. Disbursement and Audit .27 5. Project Monitoring .27 F. Project Benefits .28 G. Environmeal Aspecs ....................... ..................... 29 H. Project Risks ....30 IV. AGREFMENTS REACHED AND RECOMMVENDATION . .31 A. Agreements Reac.ed .31 B. Recommedation ................................................ 33 ANNEXES Annex I. Prioect Costs Annex IL. EsdmatBd Schedule of Bank Disbursemnts Annex mI. Organiation and Role of Coordinating Agency (FIRCO) and Project Unit Annex IV. Envionental Aspects Annex V. Selection of Project StaeS Annex VL Ecnonmc Anayis of Rural ivestumts and Soil Conservation Measur Annex VII Technology Transfer and Soil Conservation Annex VI. Project Implent Idicators, Impact Indicators and Implemnataion Schedule Annex IX Project Rexoting Annex X. Related Documents and Data Available in the Project Files MAP: ID No. 25277 'This zepot is buod onthe findia of o World Bak aial missdon which vistd Mexico fiom November 3 to 20, 1M. 'iM wision wa composed of MeasstslMme. Iohn Mocibre Crak Manr) Jobn Wse, Te.. Ron (Dan., Johbh Glenn, and 1Jo1a Woolley (consultants). The Division Chief u Midhad and dte Vaunat DrCtor is Bdeo Ssg=. TM$s dotwA has a mstricto distnona and may be used bye redpiontos ir dkte pa*= of tber. 1 k dWu tis lsoontents Synot otlmse be d siestwlbot WorldB _XiM .  MEXCO RAJNE ARMAS DEVELOPMENr PROJECT LOAN AND PROJECT SUMMARY Borrower: Nacional Financitta, S.N.C. (NAFIN) Gua_mor: United Mexican States LExcutn Agency: Trust Fund for Shared Risk (FIRCO) Beneelaries: Farmers in 10 states characterized by poor rainfed agriculture with a high dependence on corn. lIan Amount: US$85 million equivalent. Tenns: Repayment in 15 years, including a five-year grace period, at the standard variable rate. Poect Objectives: The project objective is to raise agricultural productivity in selected rainfed areas. Ihis would be done through: (a) investing in small-scale irrigation and other productive investments; (b) transferring both new and existing agricultural technologies through private sector extension; (c) reducing costs of soil erosion; and (d) strengthening the capacity of the executing agency. The proposed project is targeted to small farmers in the ten states, chiefly in semiarid and temperate zones, with rairfall less than 1,000 mm annually and where moisture deficit is the main constraint to productivity. Project Components: The proposed operation would have three components: (a) Rural Investmnt (US$147.2 million total cost) would include: (i) tubewells and related irrigation structures (US$21.7 million); (ii) other small-scale irrigation and drainage structures (US$62.1 million); (iii) soil conservation works (JS$17.7 million); (iv) other investment, including livestock production facilities and nurseries for reforestation (US$6.4 million); (v) supervision by FIRCO (US$13.5. million); and (vi) recurrent costs of operation and maintenance paid by producers (US$25.8 million). (b) A4ricudal Tecolog Transfer (US$61.0 million) is needed for the central corn-producing area of Mexico to reorient agriculture away from the traditional heavily protecd corn towaids activities that can compete under free trade. The sub-components are: (i) agricultural extension services (US$24.1 million); (ii) validation and demonstration plots (US$2.8 million); (iii) soil and water conservation practices and equipment (JS$15.5 million); and (iv) recurrent costs of soil conservation practices paid by producers US$18.6 million). All extension services and soil and wawter conservation services provided under the project would be contraced competitively through private firms. The proposed operation would  give additional emphasis to the creation of producer organizations to support agricultural extension, to the resolution of site-specific technology transfer problems, and to provision of technical assistance to producers shifting from rainfed to irrigated agriculture, or undertacing new practices, (c) - Institutional Development and TraWning (US$17.4 million) s-/ould strengthen FIRCO's: (i) capacity to design investments so as to improve rural investment sub-projects (UJS$4.4 million); (ii) staff training (US$ 3.7 million); (iii) infbrmation mainagement (US$3.6 million); (iv} expertise in the areas of soil conservation, livestock production, and monitoring and evaluation of the productivity and environmental impact of the project (US$1.6 million); and (v) project unit (US$4.1 million) Cost and Fnancing Total project costs are estimated at US$225.6 million. Project baseline costs Plan: are US$182.7 million; US$18.3 million is for physical contingencies, and US$24.6 million for price contingencies. Of the total project cost, US$178.4 million will be for investment, and US$47.1 million for incremental recurrent costs, which will be paid by benefitting producers and the Government. Benets: The project will directly benefit participating farmers in the ten selected states. ExA:ted benefits include higher production stemming from a switch from rainfed to irrigated agriculture, higher crop and livestock productivity, a more diversified crop mix, and better delivery of extension services. PRR: Subprojects must show ERRs of at least 12 percent to be eligible for inLclusion in the project. The average net present value of a sample of 26 representative projects analyzed at appraisal was estimated to be US$7,000 at a discount rate of 12 percent. Ris: Project risks are: (i) lack of counter-part finding; (ii) laci of producer contributions; (iii) lack of appropriate technical packages; (iv) faflure of farmers to continue to pay for private extension services; (v) failure to adequately operate and maintain project investments; and (vi) lack of fuids from rural banks to supply credit to participating farmers. These risks have been and will be addressed by: (i) receipt of a first-year budget commitment of Federal Government funds as a condition of negotiations and obtining budget commitments for subsequent years during the annual project reviews; (ii) requiring cost-sharing commitments by the producers before sub-project implementation; (iii) undertaking continuous review by FIRCO's technical stiff in order to ensure that unsuitable extension packages are quickly modified to meet actual requirements; (iv) competition amnong private extensionisu to provide quality services that farmers are wifling to pay for; (v) transferring ownership, operation, and maintenance of project investments to producer organizations; and (vi) estimating in the annual plan preparation each year's financial needs and discussing with commercial banks, to ensure that needs and available resources correspond. Nvaty Categoixy: Not applicable. The expected project impact on poverty will be indirect as additional income and employment fior small farmers and landless laborers  - iii - will be generatfd through incremental crop and livestock production and diversification -Ato high-value crops like fruits and vegetables. Estimated Project Costs, 1cing Plan and Disbursemtente (US$ million) Local Foreig Total L. Rural Investment 67.7 51.5 119.2 I. Agrcutural Technology Transfer 48.4 0.8 49.2 m. bnstitutional Development and Training 8.2 6.1 14.3 Total Baseline costs 124.1 $8fi6 182.7 Physical Condngencies 12.5 5.8 18.3 Price Contingencies 17.5 7.1 24.6 Total Project Costs 154.1 71.5 M25.6 IBRD 40.4 44.6 85.0 Federalgovernment 21.4 0.3 21.7 State govermments 9.0 14.7 23.7 Producers 83.3 11.9 95.2 Total 154.1 71.5 ;-25.6 Disbursemei Bank FY 1-9 1996 1997 1998 I9 Annual 23's 16 16 16 14 Cumlative 23 39 55 71 85 a/ Excluding duties and tLXes. b/ Includes an initial deposit of US$5.5 million in the Special Account and retwoactive financing of up to US$8.5 mvilon for eligible expendires incurred after November 3, 1993.  I. THE AGRICULTURAL SECTOR 1.1 The chief problems affecting Mexican agriculture are restoring growth after a long period of stagnation, man. :->. the transition from a heavily state-controlled sector to market orienta$ton, and adjusting to world prices as part of the policy goal of free trade. Solutions require further dismantling of legal and regulatory barriers to growth, a continued movement away from distorted input, output, and factor pri'mes, greater public and private investment, and more productive technology and efficient technology transfer. A. Sectoral Output and Growth 1.2 Mexican agriculture accounted for a'bout 8 percent of GDP and roughly 6 percent of total exports i 1992. While agriculture provided about 18 perceTm of the value of GDP in 1950, it fell to about 8 percent at the end of the 1980s. Productivity in the sector is so low that agriculture still accounts for about 25 percent of employment despite its small share of GDP. 1.3 Crops provide roughly 60 percent of agricultural output, livestock 30 percent, and forestry, fishing, and huuting the rest. The relative importance of crops, livestock, and forestry has shifted little since 1950. Of the crop groups, cereals account for about 45 percent of the value of agricultural output, fruits, vegetables, and legumes 29 percent, sugar cane 7 percent, oilseeds and fibers 10 percent, coffee and cocoa 6 percent, and others 3 percent. The five 'basic crops"-corn, beans, wheat, rice, and sorghum-occupied 12.3 million hectares of the roughly 13.4 million hectares in annual crops in 1992. Within the livestock subsector, cattle typically contribute a little more than half of the value of output. 1.4 Sectoral growth from 1940 to 1964 was 4.6 percent overall, 5 percent in crops, 2.6 percent in livestock, and 3.) percent in forestry. Sources of growth were expanding cropped area, shifting rainfed to irrigated land, and raising yields in some crops. Widespread adoption of new cultivars of wheat, sorghum, and oilseeds accelerated mechanization and more chemical inputs contributed to higher yields. Fertilizer use rose nearly 5000 percent from very low levels in 1950 to the mid-1980s, and about 2000 percent per hetare. Tractor use expanded on aggregate and per hectare. Capital inputs displaced labor; the latter grew only 25 percent in total between 1950 and 1985, and actually fell about 45 percent per hectare. 1.5 From 1950 to 1966, irrigated land grew by 194,000 hectares annually; from 1967 to 1985, it grew only 126,0Q hectares annually. Irrigated crops cover about one-third of cultivated land and account for half of the value of output. As real irrigation costs rose, irrigation expansion decelerated from 3.4 percent to 2. 1 percent, and, as a result, sector growth suffered. The sector grew at around 2.9 percent annuaily from 1965-1980 and at 0.4 percent annually from 1980-1990. Some of the decline in growth is further attributable to a slowdown in the expansion of cropped area, as the rate of area expansion fell from 2.8 percent in 1950 to 1965 to 0.3 percent in 1966 to 1987. 1.6 It is now much more costly to use the two traditional sources of growth: expansion of irrigation, and of cropped area. Scope for expanding total cultivated area, or increasing the irrigated areas is narrow because additional land to cultivate is of poorer quality, such as semiarid pastures, or is environmentally sensitive, such as wetlands and forests. Major expansion of irrigation is limited because of the rising cost of new gravity irrigation structures, growing competition for water from urban and industrial users, and the environmental costs of new groundwater exploitation in certain areas. Other Bank operations- such as an Irrigation and Drainage Sector Project (ll)SP; Loan 3419-ME) and a recently approved On-Farm and Minor Irrigation Networks Improvement (Project OPMIN; Loan 3704-ME)- seek to help restore the contribution of irrigated agriculture to sustainable  -2- agricultural growth. Outside the irrigation sub-sector, however, sectoral stcategy w.vli need to rely - more on measures to increase the productivity of rainfed lands, including new investments and product diversification, both of which are supported by the proposed operation. B. Strategy for a Tramition to a Market-oriented Agriculture 1.7 Past policy was highly interventionist. It set producer prices of major products well above world levels through guaranteed producer prices and trade restrictions, while subsidizing key inputs and consumer prices. A public company, National Food Supplies Company (CONASUPO), managed a wide commercial network, including milk processing plants, warehouses, urban supermarkets, and rural stores. The government supported parastatals, including 50 sugar mills, invested in irrigation and other infrastructure, restricted land transactions, mandated production targets for growers, marketed crops, and subsidized fertilizers, farm credit and crop insurance. 1.8 Those heavily interventionist policies ultimately failed to improve sector growth, rural poverty, or the nutrition of the poor. Though agriculture grew at about 4.5 percent froni World War I to 1970, it grew at less than 2 percent from 1977 through 1993. Rural poverty and malnutrition remained severe despite guaranteed prices to producers and food subsidies to cosumers. Parastatals incurred fiscal costs and economic costs in terms of market distortions. Official output targets, intended to guide production to what the government considered optimal, led to inefficient resource use. Restrictive intellectual property laws, conceived to protect Mexican consumers and to promote domestic industry, hampered technology generation and transfer. Constitutional barriers to land transactions discouraged investinmnt and diverted land into sub-optimal uses. 1.9 The past two administrations (de la Madrid, 1982-1988; Salinas de Gortari, 1988-present) have developed an agricultural strategy to achieve the transition to a market-based agricuiture. As part of the strategy to raise agricultural productivity, the Govenmment is stopping many of its historical activities in agricultural production, research, extension, investment and input use, marketing, and consumption. The strategy reduces government involvement in agriculture, eliminates nearly all trade and price protection, cuts untargeted consumer food subsidies, and seeks to target spending on public goods, poverty alleviation, and eases the move to world prices. The proposed operation fits into these objectives as it targets poor rainfed farming areas that will lose income because of the trade and price policy reforms, and directs public spending toward investments, not consumption (see para. 3.20). 1.10 Agricuitural policy reforms. The previous (1982-88) and present administrations have sought to reform agriculture as part of a general economic restructring. With, support from several Bank sectoral adjustmemt loanse both administrations have opened agriculture to intational trade, with lower and more transparent protection or taxation. They disbanded most agricultural parastatals, eliminated subsidies on fertilizer, pesticides, crop insurance, credit and machinery, slimmed the agricultural bureaucracy, removed official targets for crop production and lifed some restrictions on intellectual property. Specific reforms have been: (a) The Government removed guarantee prices for 10 of 1t commodities, except corn and beans on the grounds that they were of too great imporance politically to be taken out of the-guarante price scheme. It limited protection to wheat, barley, and oilseeds, and removed tuaation on some goods (e.g., cattle exports). It now intends to eliminate nearly all agricultural protection and to give direct income support to producers (see A Pfilizer Sector Adjusun_a Loan (Loan 2919-MB; approwed on March 15, 1988) assud in redudion of heavy fetize suWbsdies and in privatizing the ste falizer company. Two Aghicultra Sector Ad.usknet Lowas  -3 - para. 1.18). The first Agricultural Sector Adjustment Loan (AGSAL I; Loan 2918- ME) supported elimination of guarantee prices and dismantling of related trade barriers; the Second Agricultural Sector Adjustment Loan (AGSAL 11; Loan 3357- ME) supported liberalization of milk trade, and decontrolled consumer goods prices. (b) The Government liquidated many public agricultural enterprises, dismantled 17 parastatals, privatized 50 sugar mills and broke up much of CONASUPO's business. AGSAL 11 supported development of a new sugar policy that allows more efficient determination of sugar prices between growers and millers. (c) Many tariff and non-tariff barriers were reduced or completely eliminated. The share of domestic output covered by import quotas fell from 100 percent in 1984 to less than 15 percent by 1989; maximum import tariffs were cut from 100 percent to 20 percent with support from Bank Trade Policy loans (Trade Policy 1, Loan 2745-ME, in FY87 and Trade Policy II, Loan 2882-ME, in FY88). (d) Subsidies for fertilizers, pesticides, machinery, and crop insurance have been eliminated; the subsidy to electricity for agriculture has been reduced in real terms, though not completely eliminated (see para. 1.14). (e) Mandatcry public crop insurance was stopped in 1990 and the amount of subsidies from the federal treasury to crop insurance has fallen to almost nothing. Some of the subsidized agricultural credit was eliminated in 1990, though remaining subsidies are significant. A restructuring of the public rural credit portfolio, designed to separate creditworthy borrowers for private banks and to exclude the most-marginal uncreditworthy borrowers, took place in 1992. (f) Under AGSAL I and AGSAL II, the Government took steps to improve the efficiency of its food consumption programs for the poor. It increased spending on food progranis targeted to the poor, tightened eligibility to exclude wealthier consumers and began a rural food program. With support from AGSAL nI, the administration successfully implemented a pilot rural health and nutrition program. It is now continuing the reorientation of tortilla and milk subsidy programs to the poor and away from more affluent groups. 1.11 Irrigation sub-sector shategy. By recognizing that irrigation has been the motor of agricultural growth, the Government's overall strategy places particular emphasis on irrigation. The irrigation sub-sector strategy is to promote private sector involvement in new investments through the transfer of the Irrigation Districts to users' groups, to transfer costs of capital, operations and maintnance to users' groups, to modify the legal framework of the sub-sector so as to promote private investment and water markets, to target public investment to priority rehabilitation of existing works, and to limit investments in new works, in line with tight economic and environmental criteria. The Bank supports this strategy and is assisting its implementation through the IDSP and OFMIN. The proposed project is consistent with the irrigation sub-sector strategy, particularly in that it promotes private sector involvement in ownership and management of irrigation investmens and subjects those investments to strict economic and environmental criteria (see para. 3.20). 1.12 lAnd tenwe. 'Me Mexican Constitution previously prevented a special tenure class, the eji, from participating in a land markete. In 1992, the Government chadged Article 27 of the  -4- Constitution that governed land use. Major reforms included in the 1992 implementing legislation iwc. ..) an end to land redistribution; (ii) security of tenure; (iii) freedom of land allocation; (iv) freedom-to enter into economic partnerships; and (v) treedom for ejidos to change their form of associatio It is now implementing an ambitious land titling program (PROCEDE), that will free the market fo. ijdo lands and allow their use of land as loan collateral. It is expected that this change toward a land market will direct land to more efficient uses, with a major impact in the priority states for the proposed project. 1.13 Legal reforms. The Government has changed other laws affecting agriculture. A new national water Jaw (December, 1992) promotes the development of private irrigation and water markets. The Government reformed intellectual property laws affecting agriculture. The new seed law (June, 1991) allows private plant breeding, unfettered distribution of materials developed in the public research system, and freer import of commercial cultivars. The new patent law (1991) permits private ownership of food production technologies, thereby encouraging investment in their development. The new intellectual property law (1991), in an important change from earlier law, allows patenting of fertilizers, pesticides, herbicides, fungicides, foods and beverages, plant varieties; microorganisms, and biotechnological processes. 1.14 The elecrieity tariff for agricultue. A specific issue related to Banlk operations in agriculture is the low electricity tariff for groundwater, which has been a contributing factor to the excessive use of groundwater in the past. From 1990, farmers began to pay higher rates and, cons&quently, to improve water use efficiency. The Government's target was to arrive at rates equal to 100 percent of the long-run marginal cost (LRMC) of energy by 1997. Energy rates rose 440 percent in nominal terms from May 1990 to January 1993, reaching 54 percent of LRMC at that time. In February 1993, the nominal monthly increase was slowed to 0.5 percent, less than the monthly rate of inflation. On October 5, 1993, the 0.5 percent increase was slowed to zero. TMis issue was discussed extensively with the Government during preparation and negotiations of the OFMIN project. Under that operation, a study will be undertaken to evaluate the effects of the remaning electricity subsidy on groundwater use and agreement on follow-up actions will be reached before any Bank funds under that project are used for groundwater development. In the proposed project,, the groundwater development supported will be on a much smaller scale than foreseen in OPMIN and will be done on a sustainable basis (see para. ?). 1.15 Effects of the reforms. Macro-economic reforms have been broadly successful. The country bas cut inflation from more than 100 percent in 1987 to less than 10 percent in 1993. Weak growth has been re-established after negative growth and severe instability in the early 1980s. It is too soon to judge if the reforms have helped agriculture in the long-run; in the short-run, growth in the sector has not changed noticeably. Experience under AGSAL I and AGSAL n indicates that the reforms had only a mild effect on sectoral output and net agricultural exports. Private agricultural investment has remained weak and public investunent has fallen. It is expected that private investment will grow as confidence in the reforms develops, and as incentives to invest increase with the North American Free Trade Agreement (NAFTA). C. The Adjustment to NAFTA 1.16 Mexico's annual average agricultural imports from the United States were US$2.8 billion in 1989-1991; Mexican exports were US$2.5 billion, making the United States by far the largest agricultral trading partner. Mexico's agricultural imports from Canada in that period were US$0.1 billion and Mexican exports were the same to Canada. NAFTA is expected to affect the sector significanty. Te timing of NAFTA liberalization, given the present struure of trade, wiil  -~~~~~~-5 immediately benefit agriculture in Mexico more than in the United States. As Mexican protection falls more after year six of the agreement, then relative benefits wHi shift to the United Staes. 1.17 Relevant NIAFTA conditions affecting cereals, sugar and fruits and vegetables are as follows. (a) Cereals. In 1991, Mexican imports of corn, sorghum, wheat and rice from she United States totalled US$0.76 billion; Mexico exported almost no cereals to the Unit States. Existing trade barriers hold Mexican corn, whest, barley, and sorghum prices above US export parity prices. Under NAFTA, corn, the leading import, is protected by a quota for six years and subsequenty by a high tariff that falls to zero from the 7th to the 15th years of the agreement. Similar quota-tariff arrangements protet wheat, barley, and sorghum. One overall effect of NAFTA is likely to be significantly expanded exports of cereals from the US to Mexico, though little effect is expected on Mexican-Canadian agricultural trade. (b) Sugar. The agreement allows more Mexican sugar into the US market over a 15 year period, but at a cost of raising the Mexican producer price almost to the (heavily protected) level of the United States. Expected effects are to expand Mexican sugar production and exports to the United States. (c) Fruits and vegetables. Thbe agreement immediately lifts some US restrictions on Mexican exports, notably on cucumbers, watermelon, carrots, seasonal eggplant and tomatoes. Expected effects are to raise Mexican exports from irrigated areas at a significant cost in foregone com production. Increased opportunity in the US market is a significant incentive for the investments supported by the project. 1.18 The trade liberaation and direct producer income support program-PROCAMPO. As part of its adjustment to the lower crop prices that are expected with free trade, d3 Government has announced a broad policy change that will free agricultural trade almost immediately and provide direct income supports to producers during a transition period. The agricultural trade Ii' -alization component will be completed with a move to world prices fot all farm commodities by Octc jer 1, 1994. This would case the domestic producer prices of some major crops-corn, beans, wheat, barley, and sorgm-to fil quickly, though to varying degrees. For corn, the most important crop, the guarantee price of N$750/ton that held throughout 1993 will drop to N$6501ton in the fall/winter season of 1993-9d and to N$600/ton in the spring/summer season of 1994; for beans, the next most important crop in area, the corresponding changes will be from N$2,100 to N$1,800 to N$1,600. 1.19 PROCAMPO includes a direct producer income support component to compensate farmets for some of the effects of the lower free trade prices. The income support will be roughly US$110 per hectare, given to 3.3 million farmers in 1994, the transition year. Those farmers will receive a like amount in real terms for 10 years beginning on October 1, 1994. Beginning in the l1th year, the amount given will decrease linearly and become zero at the end of 15 years. 1.20 PROCAtM has severa advantages over the current system of trade protection. It does not encourage inefficient production with price incentives because it is totally delinked from the quantity of current output. It does not worsen the budget deficit by encouraging excess production of commodities that CONASUPO must then buy. PROCAMPO is more equitable than trade protection since it supports producers who received nothig from trade prtction becse they produced nothing for the markt.  U. RAINFE AGCRICULTURE 2.1 The 1991 and 1992 national agricultural statistics showed, on average, that rainfed agriculture occupied 12.1 million harvested hectares of the national total of 17.2 million hectares in annual, perennial, and seininerennial crops, excluding pastures. Irrigated agriculture occupied te remaining 5.1 million hectares. The average value of output in 1992 domestic prices was N$34.3 billion, of which 42 percent was from rainfed crops and 58 percent from irrigated. The average value of production per hectare from all crops in rainfed agriculture was about N$1148 and N$3708 in irrigated. 2.2 The main rainfed crops are corn (6.55 million hectares), beans (1 -82 million kectares), sorghum (1.06 million hectres), wheat, forage oats, and barley (ial roughly 0.25 miu1ion hectares). The productivity of those crops is much higher when they are, grown with irrigation. Rainfed crop yields were: corn, 1.76 metric tons per hectare (mt/ha); beans, 0.56; sorghum, 2.44; wheat, 1.74; forage oats, 5.38; and barley, 1.61. 'Me corresponding yields with irrigation were: com, 3.70 mt/ha; beans, 1.34; sorghum, 4.74; wheat, 4.84; forage oats, 19.23; and barley, 4.04. 2.3 Corn is crucial to rainfed agriculture as it accounts for about 30 percent of the values of rainfed output. A national survey of corn production done by the Ministry of Agriculture and Water Resources (SARH) has shown the possible effect of free trade. The -tudy indicated that 4 of the 6 million hectarus in rainfed com in Mexico would not be competitive . free trade prices. New investments in irrigation, soil conservation, and agricultural technology transfer will be required, therefore, to diversify out of rainfed corn growing areas into new activities, if extensive unemployment and sharply falling wages in those areas are not to result; the proposed project would support several such investments (see paras. 3.11 through 3.19). 2.4 The negative effects-on employment, wages, and income-of moving to free-trade prices will be heaviest in those states where agriculture is almost entirely rainfed. Of the 31 states, 15 have more than 80 percent of the cultivated area in rainfed crops. Irrigated agriculture is even more concentrated-six states have 61 percent of the national irgated area-so the benefits of free trade will accrue there, at least initially. Investments supported by the proposed project will be targeted to those states expected to lose under free trade prices based on an analysis of the current importance of rainfed corn there (see para. 3.15(a)). A. Baniers to Growth in the Rainfed Areas 2.5 The main technical barriers to agricultural growth in the rainfed areas are: (i) unreliable water supply; (ii) lack of technology transfer services; and (iii) soil erosion. Policy barriers are discue in para. 2. 10 through 2.12. 2.6 Ureliable water supply. Rainfed agriculture is fundamentally less productive than irrigated because of the lower yields due to lack of reliable and timely water supply. The principal objective of the proposed project is, therefore, to raise the productivity of rainfed lands through ifigation and soil conservation investments and related technology transfer (see para. 3.13). The project promotes technical solutions to the problem of low productivity in rainfed areas, including a shift from rainfed -toirrigated agriculture, improved efficiency of existing irigated agriculture, and invest ton;entt conserve soil and soil moisture.  -7 - 2.7 Lack of tecuology tansfer. Bank sector work" has emphasized the deficiencies in the Mexican agricultural technology transfer (extension) service. Those deficiencies have included inadequate staff training, inefficient targeting toward poor regions with small famers where private extension is weaker,- unclear focus on the problems as perceived by farmers, and lack of cost- consciousness among fte extensionists themselves. The Government is now attempting, with support from Bank loans in extension (Agriculture Training, Extension and Research Program (PROCATI, Loan 2958-ME; closed on December 31, 1992)), research and extension (Agricultural Technology Project, ATP; Loan 3465-ME) and irrigation and drainage (IDSP and OFMIN) to strengthen public agricultural extension services and to re-orient them toward small farmers. 2.8 One aspect of the Government's current extension policy is a gradual privatization of extension. Privatization of the extension services consists of replacing public extension staff with private staff, direcdy contracted by farmers, and with temporary and partial public financing of the extensionist. The proposed project supports that privatization effort (see para. 3.34). Problems to date with this scheme have been inadequate training of private extensionists, an excessive focus on individual commodities and not on overall fanm productivity, and a lack of cost consciousness. Other actions would be taken under the proposed project, which has major compcnents to improve technology transfer through private extension and to organize producer groups in support of investments (see paras. 3.22 and 3.26). The Bank-supported ATP is helping to strengthen the national agricultural research and extension system; no further action is proposed under this project to further agricultural research because FIRCO has acceptable established procedures for coordination with National nstitute for Forestry, Agriculture and Livestock Research (INIFAP) and other reseach institutions which would be maintained under the project (see para. 3.46 and Annex IlD. 2.9 Soil erosion. Erosion in rainfed agriculture is often more severe than in irrigated because it is more frequently practiced on sloping lands with poor, erodible soils. Erosion causes lower yields and higher r_oduction costs per unit of output, even though fewer inputs are used. Bank sector work and other studies have identified soil erosion as a significant problem in the project states'. For example, it is thought that 22.7 million hectares in the ten project states (see para. 3.15(b)) suffer from moderate to severe erosion; average productivity losses on lands having those degtees of erosion have been estimated to range from 37 to 75 percent of potential yield'. Under the agricultural technology transfer component of the proposed project, economically and financially profitable soil consevation measures would be extended to participating farmers. 2.10 Econoidc policy barriers. A long-standing issue in Mexican agriculture has been incentive distorticns caused by Government policies affecting land tenure, output prices and input costs, and fiacial intermediation. Recent changes in land tenure legislation, and a national land titling program (see para. 1.12) are expected to help farm investment climate in the long term; specifically, they should strengthen incentives in the project area for profitable and sustainable investment in land and water resources. Given the scope of the legal reforms affecting land tenure and the time needed for them to produce their expected effects, no further action affecting land tenure would be linked to the propoF A operation. 2.11 Government price policies provided an incentive in the past to expand the area under rainfed agriculture and to distort cropping paterns, notably to grow more corn and beans ta would be World Dank, Aexk* - AgraAft rair&teZVmdn SerW= &bsecrReporn, Report No. 8&S-ME, Noveme 19-5; and Wodd Bak, Mecw - Agrcutwal Techology Sector Review, Repodt No. 9297-MB, Augs I992. Annex X, Project files, Wold Bank, Meac - Agr*ural Sector Memo, Yeltow Cover Reot, Chopter 9, Ma 1994.  ~~~~~~~~~~-8 warranted at import parity prices. The price of coa has almost always been above world levels since 1960 and was still 70 percent above the import parity price in late 1993. The domestic price of beans bas also been well above import par 'My. The barley price is about 90 percent above impott parity and the income of wheat producers is inflated by transfers from the public treasury. The Government has announced its decision to phase out protection to the main crops by October 1, 1994 and w move to a delinkod system of direct produtcer income supports as part of PROCAMPO (see para. 1. 18); no furuther action concerning agricultural price policies would be required under the proposed project. 2.12 The rural financial system has been less effective in serving rainfed agriculture than irrigated. (Some of this was related to te land tenure system, which interfered with the use of farm land as collateral; see pama; 2.10.) High transactions costs in serving small and poor borrowers made it unprofitable for private banks to lend to them, so such borrowers were mainly reached through the public trust funds discounting loans through intermediaries. Poor administration led l latter into severe financial difficulties, and major re-structuring has taken place including impro- 4d loan administration by the banks, and re-scheduling or cancellation of overdue loans. The restructuring is believed to allow existing credit institutions to meet additional demands for credit expected within the project. No new action is contemplated under the proposed project to improve the functioning of rural financia markets. The overall rural financial system is being addressed in sector dialogue between the Bank and the Government of Mexico (GOM), including a study of rural informal lenders to identify measures to increase competitiveness and efficiency in the system. 2.13 Policies have also indirectly affected rainfed agriculture through protection to the won-agricultural sector and through the real effective exchange rate. The former harmed the competitiveness of agriculture in comparison with other domestic activities; the latter reduced the returns to producers of agricultural tradables. The broad opening of the Mexican economy that began in the mid-1980s removed much of the indirect taxation on agriculture and the situation is expected to improve furter as NAFTA is implemented. Possible overvaluation of the real effective exchange rate is a general macro-economic issue on which the Bank maintains dialogue with the Government. 1m. THE PROJECT A. Rationale for Bank Involvement 3.1 The chief rationale for public sector involvement in the proposed operation is that improving rainfed agricultural productivity requires new investment that has not been available from the private sector. Lack of private capital, caused by the poverty of the beneficiaries, has prevented investment in the past. In the near future, the shift to world prices will aggravate the previous lack of capital because the income derived from two major crops, corn and beans, will fall sharply. Moreover, the rapidity of the transition to free trade, though beneficial to consumers paying lower food prices, will harm rainfed producers who cannot adjust quickly because of the inherent short-term limitations on rainfed agriculture. The need to facilime adjustment among those producers, who have been too poor to invest in the past and whose terms of trade will fall sharply because of the policy reforms, justifies public sector support on a grant basis. 3.2 A second rationale is the public goods nature of agricultural extension and soil conservation advice. Because the information provided by agricultural extensionists is easily copied and put into practie, its supply cannot be privatized completely. Providing that information requires either direct public action or public financing of private firms or individuals that supply those services. The proposed operation supports public financing of private firms and individuals as part of the Governmen's overall sector strategy of reducing direct public intervention.  -9 - 3.3 A third rationale for public seetor involvement is the externalities related to soil erosion in rainfed areas. Those exteralities consist of erosive deposits into waterways and the permanent loss of the productive capacity of land caused by erosion. Private actions are insufficient to resolve those extemalities and hence public intervention is justified. Experience with public soil conservation efforts worldwide suggests that Goverunent intervention should be limited to the provision of information through extension "ervices (see para. 3.26), an approach that the proposed operation supports. 3.4 The primary rationale for Bank involvement is the need for additional financial resources to support incremental investments to raise agricultural productivity and to stop the expected fall in the incomes of poor farmers and landless laborers in the project staes. The GOM has specifically requested Bank support for such investnents. Though investment funds will eventually become available from 1iscal savings after elimination of the corn subsidies, and after the end of the PROCAMPO direct income support component, palliating the effects of the new policy on poor producers and landless laborers requires finds almost immediately. 3.5 The second rationale for Bank involvement is that this project, together with other operations-notably the IDSP, the ATP, the first Decentralization and Regulation Development Project (DRD ), OFMIN and the proposed DRD II and Aquaculture projects-has a catalytic role in orienting economically efficient public spending toward investments designed to alleviate poverty. Bank involvement provides an additional discipline to subprojects supported under the operation, notably in clearly defining eligible investments and a minimum rate of return. 3.6 The proposed project is part of an integrated Bank assistance strategy, as elaborated in the Country Strategy Paper of May 13, 1994 and discussed at the Board of Directors June 9, 1994. It is consistent with the Bank's strategy of promoting an increased private sector participation in the agriculture sector, guiding public investmnents to areas not well served by the private sector, and re- enforcing the role of producer groups. The IDSP assists the development, rehabilitation and maintenance of the chief irrigated areas. OFMIN promotes improvement of minor irrigaton networks and complementary on-farm structures in irrigated areas. The ATP assists with the generation of new production technology for rainfed areas. The DRD I operation supports social investments in the four poorest states. Several investment operations-Program for Integral Development of Humid Tropics (PRODERITH I and II, Loans 1553-ME and 2658-ME) and Chiapas Rural Development (Loan 2526- ME)-have promoted agricultural development in the humid tropics; sector work is now reviewing that zone to define lessons from experience and options for future strategy. B. Lessons from Bank perienc.. 3.7 The major lesson from previous Bank projects all over the world is that once the incentive framework is appropriate, it is difficult to do much for agriculture in the more arid rainfed zones beyond agriculural research, extension, and some environmental measures, such as soil and water conservation, without irrigation. Hence, the project components are limited to small-scae irrigation and drainage, livestock production, agricultural extension, soil and water conservation, and related institutional development of the executing agency (see pars. 3.17). The proposed operation targets capital imestments to zones considered to have productive potentia (see pars. 3.20); much poorer areas (f Mexico, with less productive potential, are part of Government's anti-poverty National Solidarity Program (PRONASOL), which receives Bank support under DRD 1, and are expected to obtain additional assistance under the proposed DRD U. 3.8 A second lesson is the need to involve beneficiaries closely in the design and management of the investmens. Because the investments are small and dispersed, centralized public manageme has proven to be unsu le because it is too costly. Management by the beneficiaries is necessary to - * - w i::_ 12;AAne th hAnAflciaries from desikn and operation  - 10- discourages good management and has led to the failutre of many projects. Individual investments will be prepared in close collaboration between the project executing agency and the benefitting producers with adequate financial contribution from thfe producers (see para. 3,21). 3.9 A third lesson is the importance of technical assistance to the beneficiaries. Technical -assistance, in this case in the form of agricultural extension, can improve productivity on its own and has many positive interactions with investments -in small-scale irrigation6'. Examples from Mexico include the PRODERITH. 11 and some aspects of the PRODERITH 1, PLANAT and PROCATI projects. Other examples from Latin America include the Second Brazil Agricultural Research Project (Loan 2016-BR). Project design incorporates research results from lending operations and sector work in Mexico and elsewhere. Those results include not only technical recommendations which can be made to producers benefitting from project investments, but institutional arrangementS for more efficient transfer of technology to the project area (see para. 3.34). 3.10 A fourth lesson from previous operations in Mexico has been the need for efficient infrastructure operations and maintenance. Some previous projects, such as the Small-Scale Agricultural Infrastructure operation (Loan 1643-ME), have failed to create an efficient and sustainable system for maintenance. The importance of such systems has been demonstrated in other operations, including Lower Piura Irrigation Rehabilitation 11 (Loan 1771-PE) in Peru and Cordoba II Agricultural Development (Loan 1163-CO) in Colombia. Poor maintenance has led to inefficient use of the works and to delays in new construction as the implementing agencies have slowed new works while trying to do better maintenance on the existing ones. The proposed operation requires producers to own and manage project investments, thus creating adequate and sustainable incentives for maintenance (see pars. 3.22 and 3.31). 3.11 Experience with productive projects in DRID I. The experience of productive projects in DRD I is especially relevant to the proposed operation. It was initially expected that productive projects would constitute about 30 percent of project costs of the DRD) 1. Instead, they have amounted to only 5 percent. The reasons for the shortfall are: O) poor targeting of the investments; (ii) poor investment design; (iii) lack accompanying technical assistance; and (iv) the inability of producers to meet the investment cost-sharing targets, which were 25 percent. The proposed operation has the followung ifuatures designed to eliminate those reasons for poor investment performance: Concerning poor tbrleting, the project has explicit economic and financial rate of return criteria for investment selection, and finances investments only to producer organizations that comply with FIRCO regulations designed to choose qualified organizations that can manage investments in a sustainable manner (see para. 3.21). Concerning poor design, much of the Bank's support to the proposed project will concentrate on small scale irrigation using professionally accepted designs in Mexico (see para. 3.22). Concerning lack of technical assisace, benefitting producers will contract technology transfer services under the project on a competitive basis; the operation provides additional resources for FIRCO to supervise this technical assistance (see para. 3.36). Concernig low cost sharing capacity, the Bank received assurances at appraisal that National Rural Bank (Banrural) funds could meet additional demands for credit expected under the project. Moreover, the investment and producer selection criteria under the project are expected to provide a positive and sustainable cash flow allowing producers to repay this credit. 3.12 Experience with DRD I has also shown that community grants for collectively-managed productive investments are inappropriate; the proposed project would instead rely on grants to A n:lated lesson rom Meico is the oeed to exclude credit componets from projects wich seeJk to foster technwlogy transfer, as is shown i the P,jed Performnce Audt Report, Report No. 8875 (June 29, 1990) of the Nalioa Plan for Rained Agicuur (PLANAT, so as not to bias famewrs decisions to adopt technology on the basis of sbzed credit provied by the exenson service.  - 11 - organized groups of producers who will be individually responsible for the investments (see para. 3.62). Producer organizations have proven in PRODERITH II and in PROCATI to be effective mechanisms to facilitate technology transfer and warrant further promotion under this project. C. Project Objectives 3.13 The chief project objective is to raise agricultural productivity in selected rainfed areas. This would be done through: (a) investing in small-scale irrigation and other productive investments that shift cropping patterns away from corn; (b) transferring new agricultural and existing technologies through private sector extension; (c) lessening the environmental costs of soil erosion; and (d) strengthening further the capacity of the executing agency to achieve those objectives. The project seeks to assist those farmers expected to be most affected by free-trade prices to improve their productivity, by adopting new, diversified agricultural production systems and improved soil and water conservation. The project would provide additional capital to poor farmers who will not easily make the transformation on their own, introduce the new technologies necessary for a shift to new production patterns and strengthen the ability of producers to choose profitable investments. D. Project Description 3.14 The project is targeted to the areas where rainfed corn is a major share of agricultural output (see the Map; IBRD No. 25277) because those areas are most likely to be affected severely and negatively, in employment, wages and income, by free trade in corn. Those zones typically have rainfall less than 1,000 mm annually. Annual grains, legumes and extensive ruminant livestock production are the chief activities. 3.15 Eligible States. A basic feature of project design is to direct investments to the areas where the negative effects of import-parity prices on rainfed agriculture are anticipated to be most severe. The method for targeting investments was agreed by the Bank and the Government at appraisal. The resulting final list of eligible states (Annex V) was confirmed at negotiations (see para. 4.1(a)). The method is summarized as follows: (a) The contribution of rainfed corn to the value of agricultural output--including income from all crops-was calculated for each of the 192 DDRs in the country. This showed that 92 DDRs had more than 25 percent of their income from rainfed corn and 48 had more than' 50%. (b) Ten states-Chiapas, Guanajuato, Guerrero, Hidalgo, Jalisco,-Mexico, Michoacan, Oaxaca, Puebla, and Veracruz- having 82 of the total 192 DDRs, also had 29 DDRs with at least 50 percent of their agricultural income from rainfed corn and another 19 having between 25 and 50 percent. Those were chosen as the project states. Free trade in corn is likely to have a particularly negative effect on the project states because much of their cropped area is sown to corn and their short-run alternatives are limited without irrigation. The ten states have fewer alternatives, as only 19 percent of their cropped area is irrigated, compared to 60 percent in the six states-Sonora, Sinaloa, Tamaulipas, Chihuahua, Coahuila, and Baja California Norte-that account for 50 percent of the national irrigated area. The ten states are also now less productive as their value of output per hectare is only 50 percent of that in the six irrigated states. (c) DDRs in the ten states are eligible to receive project investments. The majority of the DDRs in those states have more than 25 percent of the total value of crop output from rainfed corn, so the coverage of affected districts is high. The targeting of project investments within the ten states to the most severely affected DDRs will be an  - 12 - indicator used to monitor project implementation (see para. 3.78) and would be reviewed during the project mid-term review (see para. 3.83). 3.16 Those areas have about 6 million hectares of rainfed crop land, or about 45 percent, of tht 1991 national area in annual crops. Ii-the 48 most severely afFected DDRs, corn production accounts for over 60 percent of the rainfed areas: cultivated, and more than 45 percent of the value of agricultural production. In 44 other moderately affected districts, the figures are over 4) percent of the agricultural land, and over 20 percent of the value of production. P1. roject Components 3.17 The proposed project has three components: rural investment, technology transfer, and institutional development and training. The total project cost is estimated at US$225.6 million, including all physical and price contingencies. Rural Investment (iUS$147.2 wm;!.ion). 3.18 FIRCO's suppc.t to rural investments has five statutory objectives. These are to: (a) Improve the physical infrastructure and productivity of individual and/or groups of producers. (b) Coordinate the use of investment funds provided by the federal and state Governments as well as producers. (c) Complete and/or reactivate earlier, incomplete Government investments and thus make them productive. (d) Encourage producer-owned actions in organizing and adopting appropriate technology (thus increasing productivity), accepting the responsibility for operating and maintaining infrastructure financed with funds and, in general, increasing independence from public sector subsidies and stimuli. (e) Cooperate with producers in the establishing socially and economically adaptable soil and water conservation practices. 3.19 The investments in the rural investment component would be of two broad types. The first would be construction and rehabilitation of small irrigation schemes. The second would be reforestation, plant nursery, soil conservation, watershed protection, drainage, animal husbandry, and plant and animal disease control. These investments would allow producers to shift, at least partially, out of rainfed agriculture and into more productive farming. At negotiations, the Government agreed to a final list of eligible investment types under the project (see para. 4.1(b)). 3.20 Selection criteria for rural investments. FIRCO has well defined criteria for selection of rural investments that were accepted by the Bank during project preparation (see Annex III). The fimdamental criterion is that FIRCO concentrates on underdeveloped areas with productive potential, as de fined fom detailed agronomic and economic studies. The poorest and most marginal areas of Mexican agriculture lacking productive potential, as defined in such studies, are excluded because the productivity of investment is too low; those areas are assisted by the national anti-poverty program,  13 - PRONASOL. In addition- to being in an area of productive potential, every FIRCO investment to be financed under the proposed project must: (a) Be in a-high priority state, as previously defined (see para. 3.15). Ob) Be an eligible investment type (see para. 3.19). - (c) Meet the technical and environmental rules of FIRCO, as defined in its Operating Manual (see Annex III and Annex IV). (d) Have an acceptable economic rate of return (see para. 3.33). (e) Impose cost sharing of the initial investments by the beneficiaries in cash, labor or materials. - At appraisal, a method was agreed to calculate the economic value of the producers' contribution to the investnents, applying rural market wages as the measure of the opportunity cost of producers' labor to jobs requiring specific quantities of labor. At negotiations, the Government agreed that producers would contribute a minimum of IS percent of the investment costs of all activities finaced under the rural investment component of the project (see para. 4.1(c)). Producers in fact contribute much more than the initial 15 percent over the life of the investment because they pay for all operating and maintenance costs. In the case of livestock investments, in which producers pay for all the animal capital, producers have paid as much as 70 percent of total investnent costs. 3.21 Charact cs of beneficiaries. The producers in the target project areas have small farms, generally in the range of 3 to 20 hectares. The sample of projects analyzed at appraisad had an average of 4.3 hectares per benefitting producer. The project would target producers with fewer than 10 ha, although farmers with 10-20 ha could participate as discussed below. About 1000 rural investments would be supported under the project, and at least 280,000 people would benefit from the rural investment conponent alone. It was agreed at negotiations (see pan. 4.1(e)) that: (i) there would be no minimum farm size limit on participation in the project; (ii) producers participating in irrigated crop activities would not own farms larger than 20 hectares; (iii) projects in which any producer participating in irrigated crop activities had more than 10 hectares would be subject to prior review by the Bank and the share of such projects in the total of investments would be monitored as one of the project implementation indicators (Annex VII; (iv) the eligibility of producers participating in livestock activities with respect to farm size would continue to be detemined by the DDR Technical Committees because of the great variation in feasibility stocking rates among DDRs; (v) beneficiaries must be organized info groups for operating and managing the investment and for obtaining technical assistance; and (vi) beneficiaries must obtain technical assistance acceptable to FIRCO. 3.22 Imple ion of rural investments. The implementation of the rural investments is as follows. The Operating Rules of FIRCO (see Annex III) govern all investments. Under those rules, FIRCO coordinates the design, contracting, and execution of the works. The technical design of the investment will be done in accordance with the same professionally accepted standards in Mexico that are used in the IDSP and OFMIN. The execution of the works is done by private sector firms or individuals, with the participation of the producer organizations and the technical supervision of FIRCO (see para. 3.68 for procurement arrangements). All benefiting producers must be organized into groups for the investments. The producer organizations must sign agreements (convenios) with FIRCO in which they agree to co-finance, own, operate and maintain the invements. (Land within projea areas is held individually and producers are free to manage their own plot as they see fit.) Upon completiot, ownership of the investment is transfefred to the producer organation which  - 14 - 3.23 The rural investment component has five sub-components: (i) wells and associated irrigation facilities; (ii) other small-scale irrigation and drainage; (iii) soil conservation works; (iv) other agricultural investments; and (v) supervision of works by FIRCO. 3.24 Wells and associated irrigation faclities (US$21.7 miUlion). The sub-component would finance groundwater irrigation from tubewellsf, specifically well digging, electric pumps, irrigation canals, dikes, drainage structures and related facilities. The project would finance about 333 wells in the I0 states, or about one well per DDR every year. The average cost of a well is about US$85,000, including drilling, lining, electrification and putchase of pumps. The water distribution' - system cost is more variable depending- upon the type of distribution (canals, tubes), the distance water is transported and type of irrigation used. FIRCO also finances many irrigation Tehabilitation projects frequently where wells have been drilled but not equipped and water distribution systems not completed or where earth canals are replaced with concrete lining. In the later case, there is a much more rapid and efficient use of water. All well digging requires prior verification by the CNA that the groundwater extraction is sustainable and that the well is not in an exclusion zone (zona de veda; see para. 3.88) 3.25 Other smnll-scale irrigation and drainage (US$62.1 million). This sub-component includes the construction of small catchment reservoirs, small surface irrigation systems usually less than 80 hectares and the extension or improvement of existing irrigation systems. The project would finance about 505 such investments in the 10 states in five years, or about 1.2 projects per DDR every year. Project investments would typically consist of: (i) electrical pumps; (ii) electric lines; (iii) irrigation ditches; (iv) PVC pipe; and (v) installation of water management structures. These are high-priority investments for FIRCO, state Governments and producers. Water is obtained from surface or ground sources and is distributed though open canals or PVC tubing. Mostly irrigation is by gravity flow but there is increased emphasis on sprinkler or other water-conserving systems. For grain crops and forage production irrigation is sometimes used as a supplement to rainfall. Where subsurface water is used, drilled wells are quite deep (200 m), although shallow wells are possible in some areas with relatively rapid recharge of the aquifer. Electric pumps are normally used to lift either surface or subsurface water to the canal or tube levels7'. Another type of irrigation investment is small earth dams to retain rain water for uss in the dry season. These are multi-ue ponds that help to recharge the aquifers, thus increasing the water availability in downstream shallow wells and providing water for livestock, allowing for very limited irrigation and slowing down the rate of rain water flow, thus decreasing gully formation. 3.26 Soil conservation works (US$17.7 miDlion). Effective and economically profitable soil conservation measures exist, but producers do not always use them because they lack information or because they are not financially profitable to producers without the attractive financing by the project. A major sub-component of the proposed project, therefore, is promotion of soil conservation. 3.27 There are different categories of soil and water conservation activities, including small earth dams (see para. 3.25), construction of various obstacles to water flow and erosion in ravines, and terracing of arable lands. The ravine control mechanisms are effective in the long term only if accompanied by revegetation of their slopes and bordering areas as well as changes in cultural practices in adjoining arable areas, which would be supported by the agriclttural extension sub- component. Physical works to improve soil conservation would be done through the rural investment component by promotion of watershed protection and soil conservation works; a separate sub- component of the agricultural technology transfer component would address cropping practices needed to promote bet soil conservation. The project- would support about 2500 activities of this type.  - 15 - 3.28 The soil conservation nmeasures assisted under the operation would be designed according to SARHI guidelines8 and the experience in other Bank projects, notably PRODERITh 11 and PROCATI. Emphasis would be given to on-farm conservation practices, such as green manuring, vegetative barriers and minimum to zero tillage, rather than to physical works requiring high initial investments and costly maintenance (see Anmex VII). Bank experience in soil and water conservation" suggests that project emphasis should bebput on agronomic practices, not on the construction of works imposed by a central authority. The former has been shown to be more profitable in the short-run and more sustainable in the long-run because it elicits greater farmer- participation in the maintenance of the works. Soil conservation activities supported by the proposed project include both practices and works; the latter are designed, constructed and maintained with significant input from producers, so as to avoid incentive problems. $oil conservation works-including terracing, banks and ditches, and small-scale reforestation-would be used in place of cropping practices on steeper slopes and in marginal areas where practices are less profitable. All soil conservation activities to be financed under the proposed project, both cropping practices and works, would be financially profitable to the farmer and economically profitable to society, effective in reducing soil losses, and envirownentally sustainable. 3.29 Other agricultural investments (US$6.4 million). FIRCO supports other small investments to intensify resource use, or to transform a particular agricultural acdvity into more productive ones. The operation would support approximately 170 such investments. The principal subprojecs are for livestock production-installations such as fencing, dipping tanks, corrals, pastures, and feed stores. This class of activity includes investments that are of special concern to women, such as small-scale (traspado) livestock production. In some instances, the participating producers already own the livestock; in others, they would buy them during project implementation. Costs of acquiring livestock are not included in project costs and are financed entirely by producers; they would not be eligible for Bank financing. The principal livestock-related investments are pasture establishment, silo construction and providing sanitary sheds or buildings for dairy cattle and pigs, water, fencing and dipping vats. Producers also provide animals for the livestock-related investments; the animals are not included in project costs because it is difficult to verify their existence and unit prices. 3.30 The chief purpose of the investments in wells, associated ir.-Igation facilities, and other small- scale irrigation and drainage is to shift the cropping pattern away from rainfed corn into crops that are profitable at international prices. Such crops include fruits, vegetables, pastures, tree crops, and, where high yields are possible, grains and oilseeds. Annex VI presents a representative sample of 33 crop production investments used for economic analysis. Of those 33 investments, the following five types are domiant Particular investment types would be selected for different locations within the 10 project states according to climate, the availability of ground or surface water, existing irrigation facilities, soil type and market access. (a) Type I - Grains, ruits and vegetables. A major type of investment would assist diversification of cropping patterns out of grains and legumes, which are familiar to participating farmers, to include fruits and vegetables. Additional income would accrue to producers from fruits and vegetables, which typically have higher yields and higher unit prices. Yields of grains would also increase with irrigation investments. The average gross income would increase from roughly N$1,250/hectare with a Sc Anrnex X, Ptojept Ales, SARHA, Gum Tomca pmCos on en ci Ttopo. T AsiA e%pv i reptd in Joi B. Dooles an WWim W Magah ds.), gersed Developmenn - _ *-- . U0..._16 OI*_4.nLA1 P,r Number 12.7 un. 19O). Tbe Mei  - 16 - cropping pattern of grains alone to N$7,360/hectare with a cropping pattern having tomato, squash, strawberry and lettuce. (b) Type I- Pastures, grains, legunes, frits, and vegetables. A second type includes pasture crops, which are expected to become more valuable as incomes rise in Mexico and which can attain very high and profitable yields with irrigation. The value of output per hectare in this type would rise from N$6,400 to N$8,400 per hectare after the investment is made. (c) Type IS - Tree crops. A third type is a mix of high-value tree crops. Tis type is technically the most complex, but offers the possibility of high profits. One investment based on tree crops provides technical assistance and land improvements to existing plantations of banana, lemon, mango, coconut and guava produced by farmers who also grow sorghum and corn. A second sub-type includes orange, lemon,- tangerine and mango grown by farmers with tubewell irrigation that also permits the cultivation of tomato, watermelon and corn. (d) Type IV - Grain and legumes. A fourth type which includes grain production-notably corn, wheat, sorghum, and rice-can be profitable even at world prices, if irrigation is available and adequate technical assistance is provided to growers. A major type of project investment is small-scale irrigation, from surface or ground sources, to produce grains for the market. The effect of irrigation, more fertilizer use and technical assistance to teach farmers how to manage more intensive production systems is expected to improve corn yields from about 2 mt/ba to as much as 6 mt/ha, bean yields from 1.2 to 2.5 mt/ha and sorghum yields from 2 to 5 mt/ha. (e) Type V - Ranfed crops. A final type includes various rainfed crops, mainlv cereals, legumes, pastures, and tree crops. The principal external input into this farm type would be agricultural extension services and, in some cases, an increased focus on soil and water conservation. 3.31 Recurrent cost for operation and intenace of rural investments (US$25.8 million). Operation and maintenance of rural investments has been a problem with many Bank-supported projects in Mexico (see pares. 3.10). The proposed project solves this problem by giving full ownership to the benefitting producers and by leaving them all the responsibility for operating and maintaining the investment. This is now FlRCO's practice, and, while it does pose risks in terms of the producers' commitment to the investment over its full life, those risks are acceptable in light of the superior incentives for good operation and maintenance offered by the practice. All recurrent costs of operation and maintenance under the rural investment component would be paid by producers' organizations, including those incurred for soil conservation works. 3.32 Supervision of works (VS$13.5 million). Technical supervision by FIRCO is vital for successful initiation of each investnent and would be required for each investment supported under the proposed project. Supervision would consist of field visits and interviews with beneficiaries, extensionists, and contractors. FIRCO now provides such technical supervision to rural investments and their associated technology transfer activities from project design through compIetioni. FIRCO has adequate staff and capacity for supervision and the operation would provide only incremental financingin this area. The incremenal costs of supervision were estimated at appraisal to be roughly '