Research Report

International Water Management Institute

INTERNATIONAL WATER MANAGEMENT INSTITUTEP O Box 2075, Colombo, Sri Lanka

Tel (94-1) 867404 • Fax (94-1) 866854 • E-mail iimi@cgiar.orgInternet Home Page http://www.cgiar.org/iimi

ISSN 1026-0862ISBN 92-9090-364-3

25

Impacts of Colombia’s CurrentIrrigation Management TransferProgram

Douglas L. VermillionandCarlos Garc s-Restrepoé

Research Reports

IWMI’s mission is to foster and support sustainable increases in the productivity of irrigatedagriculture within the overall context of the water basin. In serving this mission, IWMI con-centrates on the integration of policies, technologies and management systems to achieveworkable solutions to real problems—practical, relevant results in the field of irrigation andwater resources.

The publications in this series cover a wide range of subjects—from computer modelingto experience with water users’ associations—and vary in content from directly applicableresearch to more basic studies, on which applied work ultimately depends. Some researchreports are narrowly focused, analytical, and detailed empirical studies; others are wide-rang-ing and synthetic overviews of generic problems.

Although most of the reports are published by IWMI staff and their collaborators, wewelcome contributions from others. Each report is reviewed internally by IWMI’s own staffand Fellows, and by external reviewers. The reports are published and distributed both inhard copy and electronically (http://www. cgiar.org/iimi) and where possible all data andanalyses will be available as separate downloadable files. Reports may be copied freely andcited with due acknowledgment.

1iii

Research Report 25

Impacts of Colombia' s CurrentIrrigation Management TransferProgram

Douglas L. VermillionandCarlos Garcés-Restrepo

International Water Management InstituteP O Box 2075, Colombo, Sri Lanka

The authors: Douglas L. Vermillion and Carlos Garcés-Restrepo are Senior Social Scien-tist and Senior Irrigation Specialist, respectively, at the International Water ManagementInstitute.

IWMI gratefully acknowledges the financial support for this study, from the Inter-Ameri-can Development Bank (IDB), the Ford Foundation, and the Federal Ministry for Eco-nomic Cooperation and Development (BMZ), Germany. The authors wish to thankEngs. Luis A. Mora and Claudia Alvarez, and Soc. Alba L. Giraldo, former IWMI-Colombia national staff, who were instrumental in field wok and data collection.

Vermillion, D. L., and C. Garcés-Restrepo 1998. Impacts of Colombia’s current irrigation man-agement transfer program. Research Report 25. Colombo, Sri Lanka: International WaterManagement Institute.

/ privatization / irrigation management/ irrigated farming / policy / costs / economic aspects /operation / maintenance / agricultural production / Colombia /

ISBN 92-9090-364-3ISSN 1026-0862

© IWMI, 1998. All rights reserved.

Responsibility for the contents of this publication rests with the authors.

The International Irrigation Management Institute, one of sixteen centers supported bythe Consultative Group on International Agricultural Research (CGIAR), was incorpo-rated by an Act of Parliament in Sri Lanka. The Act is currently under amendment toread as International Water Management Institute (IWMI).

3iiiiii

Contents

Summary v

Introduction 1

Methodology and Study Sites 2

Management Transfer Program in Colombia 3

Transfer Process in Five Districts 5

Impacts of Transfer 13

Conclusion 32

Annex 1. Map of RUT District 35

Annex 2. Background Information on the Five Sample Districts 35

Literature Cited 38

v

Summary

This report deals with the results of a study con-ducted by the International Water ManagementInstitute (IWMI) in collaboration with the NationalInstitute for Land Development (INAT) to assessthe impacts of the current national irrigation man-agement transfer program in Colombia. In 1990,the Government of Colombia adopted a new na-tional policy to transfer management of its irriga-tion districts to water users’ associations (WUAs).This report examines the context of transfer, thebasic transfer strategy, powers and functions de-volved, and the impacts of transfer on irrigationmanagement and irrigated agriculture in threesample irrigation districts.

Two districts that were not transferred until1995 and 1996 were also selected to enable com-parison between schemes that had and had not yetbeen transferred during the period of historicalanalysis. Impacts measured include costs of irriga-tion to the government and the farmers, financialsolvency of the irrigation districts, quality of irriga-tion operations and maintenance, and the agricul-tural and economic productivity of the irrigationsystems. Data were collected through interviewswith irrigation management staff and farmers, sec-ondary data available from irrigation and agricul-ture departments, a sample survey of farmers, anddirect inspection of irrigation networks.

The national irrigation management transferprogram in Colombia adopted in 1990 can be char-acterized as significant but only as a partial devo-lution of management to water users. The govern-ment maintained considerable advisory influenceover the districts for several years, exercising some

control over O&M plans and budgets, and resist-ing district attempts to release large numbers ofstaff. After adoption of the 1993 Land Develop-ment Law, in 1994, this control has been relaxedconsiderably as districts gained almost completecontrol over management. However, powers de-volved do not include a formal water right or own-ership of irrigation scheme infrastructure. Also, thegovernment has not made it clear whose responsi-bility it will be, and under what terms and condi-tions to finance possible future costs of rehabilita-tion.

Management transfer prompted a number ofmanagerial changes aimed at improving manage-ment efficiency and accountability of district staff.Transfer resulted in a significant shift in the bur-den of cost from the government to farmers, whichhas generally been accepted by farmers. But trans-fer has not had substantial impacts on the perfor-mance of operations and maintenance, or on theagricultural and economic productivity of irrigatedland or water, neither improving negative perfor-mance nor causing detriment where performanceis positive.

This report raises concerns about how lack ofa comprehensive devolution policy for the irriga-tion subsector can discourage farmers from invest-ing in the long-term sustainability of their irrigationschemes. More attention is needed toward usingthe transfer process to create local managementself-reliance and ensuring that needed support sys-tems for local management are in place prior toimplementation of irrigation management transferprograms.

1

This report deals with results of a study toassess the impacts of the current nationalirrigation management transfer program inColombia. The report examines the contextof transfer, the basic transfer strategy, andits impacts on selected systems. Impactsmeasured include costs of irrigation to thegovernment and the farmers, financial sol-vency of the irrigation districts, quality ofirrigation operations and maintenance,and the agricultural and economic produc-tivity of the irrigation systems.

In 1976, the Government of Colombiafirst transferred management of irrigationdistricts from the government to WUAs forthe Coello and Saldaña irrigation districtsin the Tolima valley in central Colombia(Vermillion and Garcés-Restrepo 1996). Af-ter a hiatus of a decade and a half, largelydue to economic recession and politicalconcerns, in 1990, the government adopteda new national policy to transfer manage-ment of all of its irrigation districts toWUAs. In the same year, the governmentresumed transfer of irrigation districts, be-ginning with the transfer of theRoldanillo-Unión-Toro (RUT) and RioRecio districts, in the relatively prosperousCauca and Tolima valleys.

Global Trend toward Devolution

Largely driven by government fiscal short-ages and a common inability to raise suf-ficient revenues from collection of watercharges, since the 1970s, governmentsaround the world have adopted programsone after another to devolve responsibilityfor irrigation management to WUAs(Johnson, Vermillion, and Sagardoy, eds.1995). Consistent with overall structuraladjustment programs, irrigation manage-ment transfer has been supported by themajor international development banks(World Bank 1993; EDI 1996; Arriëns et al.1996). The reforms generally include effortsto organize WUAs, train future managers,make essential structural repairs, and ne-gotiate and formalize agreements betweenthe government and the water users.

As the reforms are normally motivatedby financial pressures and driven by do-nor deadlines, devolution policies tend tobe adopted ahead of identification of apractical and integrated strategy for imple-mentation. There is a significant gap ofknowledge about actual results of IMT—which strategies do and do not work, andwhat prerequisites are necessary to sup-

Impacts of Colombia's Current Irrigation ManagementTransfer Program

Douglas L. Vermillion and Carlos Garcés-Restrepo

Introduction

2

port sustainable local management of irri-gation. These questions are global in rel-evance. To date, there has been a lack ofsystematic, comparative research to an-swer these questions (Vermillion 1997) tofacilitate policy makers and farmer leadersto make better strategic decisions and in-vestments. This report is part of a com-parative research program at the Interna-tional Water Management Institute to ex-amine the modalities and impacts of IMTin several countries through a commonmethodology (Vermillion et al. 1996).

The most important research ques-tions related to management transfer are—

1. Under what conditions should trans-fer be attempted or not?

2. What kinds of external support arenecessary to make transfer work?

3. What have been the impacts of transferin those districts already transferred?

The third issue is the focus of this report,although some of the findings herein contrib-ute partially to the other two questions.

1In all cases, sample siz-es were more than 5 per-cent of farmers. Samplesizes were 91 farmers inRUT, 59 in Rio Recio, 88in Samacá, 56 in SanRafael, and 90 in MariaLa Baja.2Some data were un-available for Rio Recioand Maria la Bajaschemes, hence they arenot included in some ta-bles and charts.

Methodology and Study Sites

This report first describes basic aspects ofthe agricultural, socioeconomic, andphysio-technical context of selected irriga-tion systems in Colombia. Then it describesthe policy and basic arrangements formanagement transfer. Against this back-drop, impacts are measured and to an ex-tent, explained. Data were collectedthrough interviews with irrigation man-agement staff and farmers, secondary dataavailable from irrigation and agriculturedepartments, a sample survey of farmers,and direct inspection of irrigation net-works.1 Sample farms were selected byrandom stratified sampling, in head,middle, and tail locations along canals.Annex 1 is a map of the RUT District

showing locations of sample farms thatwere thus selected (Giraldo 1997).

The report is a comparison of trendsin performance in five schemes. Three ofthe irrigation districts selected for thestudy (RUT, Rio Recio, and Samacá) weretransferred early enough to enable com-parisons of performance for 4 or 5 yearsbefore and after transfer.2 The remainingtwo schemes—San Rafael and Maria LaBaja—enable comparison between alreadytransferred and not yet transferredschemes for the period of analysis between1986 and 1995, as the former was trans-ferred at the end of 1995 and the latter in1996.

3

Irrigated Agriculture and TransferPolicy

Approximately 80 percent of Colombia hasa humid and tropical, or warm temperateclimate. Of an estimated 18.3 million hect-ares of cultivable land, 3.8 million hectaresare actually cultivated at present. Of anestimated 7.4 million hectares of land thatis potentially irrigable or could be produc-tive with installation of drainage or floodprotection facilities, 750,000 hectares areactually equipped with irrigation or drain-age facilities at present (MOA 1990). Co-lombia still has considerable potential toexpand the area under irrigation.Colombia’s mountainous terrain, highrainfall, and fertile volcanic soils createconditions conducive to small-scale devel-opment of irrigated agriculture. A widerange of crops, from tropical to temperate,can be grown. The most widely cultivatedcrops are rice, maize, sorghum, soybean,pasture, potato, and vegetables.

The private sector has developed andmanaged 463,000 hectares (or 62%) of thecurrent functional irrigable land in Colom-bia. The public sector has developed only38 percent. Development and operatingcosts of public irrigation schemes are onaverage about double those of privatelydeveloped schemes (FAO 1994).Colombia’s extensive experience with irri-gation development in the private sectorand the sector’s well-known higher levelsof economic efficiency have created generalexpectations that farmers, given the rightcircumstances and training, are capable oftaking over management of public irriga-tion schemes and managing them more ef-ficiently than the government.

In Colombia, the first impetus for man-agement transfer came from water usersthemselves, who lobbied the governmentto take over management of the schemes inthe Coello and Saldaña districts in theTolima valley. In response, the governmentcreated HIMAT3 in 1975 and gave it(among other responsibilities) the dualmandate to first, take over management ofall 23 public irrigation districts in Colom-bia and second, establish and form WUAsin the districts to eventually take overmanagement from HIMAT. Coello andSaldaña districts were transferred to theusers in 1976 under an inter-sectoral con-stitutional principle referred to as “delega-tion of administration” (Plusquellec 1989).However, the delegation did not includetransfer of ownership for scheme assetsnor full WUA control over budgets, O&Mplans, and personnel. HIMAT retained astrong supervisory role in administeringbudgets and O&M plans, and strongly re-sisted early attempts by the WUAs to re-duce the number of staff employed by thedistricts (who were HIMAT staff before thetransfer).

Nevertheless, the relatively successfulresults of this experiment (Vermillion andGarcés-Restrepo 1996) strengthened politi-cal resolve to make management transfer anational policy. Toward the end of the eco-nomic recession of the 1980s, the govern-ment was ready to resume implementationof the devolution policy as part of its over-all strategy of economic liberalization andpolitical decentralization.4 Roldanillo-Unión-Toro (RUT) and Rio Recio were thefirst irrigation districts to be transferred aspart of the “new wave” of transfers, in1990. Transfer of several other districts fol-

Management Transfer Program in Colombia

3HIMAT is the Institutefor Hydrology, Meteo-rology, and Land Devel-opment. In 1994, respon-sibility for meteorologywas removed from theagency and its namewas changed to the Na-tional Institute for LandDevelopment (INAT) toreflect its narrower fo-cus on development ofirrigation, drainage,and flood control facil-ities.4The strategy includedremoval of agriculturalprice supports, inputsubsidies, and tradebarriers.

4

lowed over the next few years (Garcés-Restrepo and Vermillion 1995).

Recognizing the problems inherent inthe partial delegation of administration,and needing to induce greater farmer in-vestment in future expansion efforts, thegovernment passed the Land DevelopmentLaw No. 41 in 1993 and its associated en-abling Decree Nos. 1278 and 2135. To-gether, the new law and decrees deter-mined that transfer would thereafter placefull control over irrigation district finances,O&M procedures, and personnel in thehands of the WUAs. The new agreementswere referred to as “concessional con-tracts” rather than delegation of adminis-tration. This was a significant enlarge-ment of devolution. However, even underthe new law, the WUAs were only givenuse rights, not ownership over irrigationinfrastructure.

Program Implementation

Under the transfer process, HIMAT, or,INAT, as it was renamed after 1994, facili-tated the formation of a WUA. This in-cluded preparation of a constitution, for-mulation of bylaws, and designation ofbasic rules and sanctions. Farmers electedrepresentatives to a Board of Directors.This was followed by preparation andsigning of a concessional contract betweenINAT and the WUA. In each scheme, thetransfer process may or may not includetraining, rehabilitation, or changes in

O&M procedures, water fees, or personnel.All such issues are resolved on a case-by-case basis through negotiations betweenthe WUA and INAT. At the beginning of1990, only Coello and Saldaña (with a to-tal irrigable area of 39,603 ha) had beentransferred. By the end of 1996, 17 of the23 public irrigation districts in the countryhad been transferred to WUAs. This con-stitutes 115,695 hectares of a total of241,077 hectares of irrigated area previ-ously under public management (Alvarezand Garcés-Restrepo 1996).

Originally all districts in the countrywere planned to be transferred by 1997. Thegovernment clearly pursued a strategy oftransferring management of the “easier”districts first. These tended to be districtsthat were more prosperous and financiallyviable, which did not have major infrastruc-ture disrepair and where farmers were morewilling to take over management. However,by 1997, after experiencing difficulties intransferring Maria La Baja and other dis-tricts, the government temporarily discon-tinued implementing the program, ostensi-bly to enable it to conduct more in-depthanalyses of what to do about the remainingdistricts considered to be most problematic.The remaining districts tended to have vari-ous problems such as high costs of irriga-tion, technical problems, facilities in a stateof disrepair, lower productivity, lower prof-itability of irrigated agriculture, social un-rest, poverty, and inability to collect ad-equate irrigation fees from farmers.

5

Transfer Process in Five Districts

The location of the five sample districts isindicated in figure 1 and basic informationon each district is summarized in tables 1and 2, and annex 2. These are small tomedium-size schemes, ranging from anirrigated area of 560 hectares to 10,200hectares and with design discharges from0.6 m3/s to 14 m3/s. Out of these fiveschemes, two are river lift pump schemes,two have small reservoirs and the other is

a river diversion scheme. Annual rainfallvaries from 690 mm in the Samacá schemein the mountainous Boyaca State to 1,890mm in the Maria La Baja scheme on theAtlantic coast (table 1). As indicated intable 2, except for the Rio Recio district,the large majority of farm holdings are lessthan 10 hectares each. In Samacá and SanRafael, 95 percent of farm holdings are lessthan 5 hectares each in size.

FIGURE 1.Map of Colombia showing locations of sample irrigation districts.

6

TABLE 1.Basic characteristics of sample districts.

Item RUT Rio Recio Samacá San Rafael Maria LaBaja

State Valle Tolima Boyacá Boyacá Bolivar

Design area (ha) 13,000 23,600 3,000 590 19,600

Irrigated area (ha) 9,700 10,200 2,893 560 9,260

Water source River lift River diversion Reservoir River lift Reservoir

Intake structure Pump Gated weir Vertical gates Pump Verticalgates

Designed Q (m3/s) 14 11 1 0.6 20

Main canal length (km) 87.7 38.7 29.7 Buried pipe 58

Total canal length (km) 170.7 135.8 58 Buried pipe 284.4

Area per km canal (ha) 57 74 51 – 33

Turnout type Pump Sliding gate Pump and Riser Sliding gatesliding gate

Control structures (#) 16 234 69 16 na

Lowest level water measured Along main Farm inlets Main intake Main intake Main intakecanal

Water delivery efficiency (%)* 53.7 74.1 86.1 80.5 54.7

Main soil type Clay, loam Clay, loam Clay, loam Clay, clay-loam Loam, clay-loam

Average annual rainfall (mm) 1100 1300 690 783 1890

Main crops Cotton, Rice, Onion, Pasture, Rice,grape, fruit sorghum, potato, pea vegetables sorghum,

trees cotton pasture

Year of transfer 1958-70 1949-51 1945 1970 1962-65

Period or year of construction 1990 1990 1992 1995 1996

WUA Asorut Asorecio Asusa Asochicamocha Asodimar

Heavy equipment (#) 27 20 2 6 26

*Water delivery efficiency is the ratio of volume delivered to volume diverted.

TABLE 2.Farm size distribution in sample districts.

Percentage of farm holdings in each size range

Farm size range (ha) RUT Rio Recio Samacá San Rafael Maria La Baja

<5 75 14 95 95 46

5.1—10 11 11 3 2 29

10.1—20 7 23 1 2 15

20.1—50 6 23 1 1 7

>50 1 29 0 0 3

7

The Five Sample Districts

Roldanillo-La Union-Toro (RUT)

The Roldanillo-La Union-Toro (RUT) irri-gation district is located in the prosperousCauca valley and serves 9,700 hectares.RUT was the first district to be transferredin the national transfer program in Janu-ary 1990. The WUA was organized at thetime of construction of the project, butfunctioned only in an advisory capacityuntil transfer, after which it became a gov-erning body (table 3).5 As part of thegovernment’s overall policy to eliminatesubsidies to the agriculture sector, the gov-

ernment halted its expenditures for O&Min the scheme after transfer. Before transfer,the subsidy was approximately 60 to 80percent of total costs. Since the schemehad been rehabilitated before transfer, noarrangements were made for further re-pairs as part of the transfer process, exceptfor some minor repairs to the main canal.After transfer, farmers began to realize thatthey had seriously underestimated pump-ing costs without a subsidy. They havesince pressured the government to providea temporary subsidy of approximatelyUS$800,000 for energy costs. No trainingwas provided as part of the transferprocess.

5Information for tables3, 4, and 5 was obtainedfrom key informantsamong district boardmembers and manage-ment staff, as well asfrom INAT officials.

TABLE 3.Occurrence of transfer activities in sample districts.

Transfer activity RUT Rio Recio Samacá San Rafael Maria LaBaja

Establish WUA Yes Yes Yes Yes Yes

Train stet farmer representatives No No No Limited Limited

Train stet management staff No No No Yes Limited

Revise O&M procedures and/or No No No Yes Yesplans

Revise water charges Yes Yes Yes Yes Yes

Reduce or eliminate govt. financing Fully Fully Fully Partly Partly

Remove some government staff Yes Yes Yes Yes Yes

Repair or improve intake and/or Yes Yes Yes Yes Yesmain canals (major) (major) (after IMT)

Repair or improvement of No Yes Yes No Yessubsidiary network (major) (major)

Did farmers help prioritize No No No Limited Limitedimprovements?

Farmer investment in No No No No Noimprovements?

Future responsibility for rehabilitation No No No Yes Yesknown?

8

Rio Recio

The Rio Recio district is located in theTolima valley in central Colombia. It irri-gates 10,200 hectares of farmland. RioRecio was transferred to the WUA(ASORECIO) in January 1990, under theprinciple of delegation of administrationwherein the government agency HIMATcontinued to exercise guidance over thedistrict for budgets, O&M plans, and per-sonnel.6 The water users favored transfermainly because they perceived that someof the funds from the water charge werebeing diverted to HIMAT budgets at stateand central levels. They felt they couldmanage the system at a cost lower thanthat of the government. As part of the ne-gotiations for a transfer agreement, the us-ers successfully lobbied to reduce thenumber of staff after transfer and avoid re-paying the cost of rehabilitation done priorto transfer (which cost the governmentUS$3.6 million). Farmers had not partici-pated in the rehabilitation and resisted at-tempts by the government to recover thecosts from them. However, several mem-bers of the staff were retained by the WUAafter transfer. Consequently, there was notraining or change in basic O&M proce-dures associated with the transfer (table 3).

Samacá

The Samacá irrigation district is located inBoyaca State, northeast of Bogota. It has aservice area of 3,000 hectares. In Samacá,negotiations for transfer began in 1991and culminated in October 1992 with theofficial transfer. As part of a strategy tomake the district financially self-reliant bythe time of transfer, in 1991, the govern-ment raised the fixed area water fee by 170

percent while abolishing the volumetric fee(mainly due to difficulties of measuringwater delivered to the farm level). Thefarmers’ main interests in transfer were toreduce management costs and improve re-sponsiveness of district management staffto the farmers’ diverse irrigation needs.Minor repairs were made in the main ca-nal and subsidiary network prior to trans-fer, but farmers did not participate in theimprovements (table 3). Government ex-penditures for O&M were discontinued af-ter transfer.

San Rafael

San Rafael district is a small river liftscheme built in 1970. It irrigates 560 hect-ares of land located in the mountainousBoyaca State. Recently, the district hasbeen incorporated into the largerChicamocha irrigation district that is cur-rently under development. San Rafael isthe first, and still the only unit in thelarger district to be functional. The WUAfor the larger Chicamocha district was or-ganized in 1995, and at the end of theyear the district as a whole (including SanRafael) was officially transferred to theWUA. As part of the transfer negotiations,INAT agreed to only gradually discontinuesubsidies over 3 to 5 years after transfer.At the request of farmer representatives,INAT also agreed to repair the pump sta-tion and to transfer farm and heavy equip-ment to the WUA at no cost to the waterusers. The WUA also successfully lobbiedfor the right to keep revenues obtainedfrom the sale of water to a beer factory andtourist center nearby. Training in pumpoperation was given to the WUA duringthe transfer process.

6As with RUT, this over-sight role continued un-til after the enactment ofthe Land DevelopmentLaw of 1993.

9

Maria La Baja

The Maria La Baja irrigation district is lo-cated 50 km south of Cartagena City nearthe Atlantic coast. Its current service areais 9,260 hectares. In June 1994, INAT andthe WUA for the Maria La Baja district(ASODIMAR) signed a consessional con-tract. However, due to reluctance by waterusers and INAT, the actual transfer didnot occur until the end of 1996. Farmerscomplained that many of the structureswere dilapidated and should be repairedbefore the system was handed over to theWUA. INAT had reservations about theperceived weak managerial capacity of theWUA. Hence, it was agreed that INATwould continue to jointly manage the dis-trict with the WUA for at least 6 monthsafter the transfer, with a possible exten-sion.

This period of joint management wasextended to a full year, and was imple-mented in 1997, when extensive trainingwas provided to district staff. Followingthe period of joint management, the WUAwould take over full control of watercharges, budgeting, and O&M plans. Itwas also agreed that a modest amount ofrehabilitation of broken-down structureswould be done after transfer and that theWUA would have a role in prioritizingwhat should be repaired. However, whichparty will be responsible to finance thecosts of future rehabilitation was not clari-fied.

Powers Devolved, FunctionsTransferred

Table 4 summarizes powers devolved andfunctions transferred to WUAs in thesample districts. In all cases, the WUA isa legally recognized entity with authority

to manage water diverted to a defined ser-vice area. The WUA is a nonprofit, quasi-municipal entity with legal rights of wayfor irrigation canals and structures. Theassociation has the right to select its lead-ers, make rules, and impose sanctions—upto the maximum penalty of fines for dam-ages to property, cessation of the water ser-vice, and taking the violator to court.

Each WUA consists of a general assem-bly of members and an elected board of di-rectors. WUA members are all landownersholding land within the command area,whether irrigated or not. A general assem-bly of members meets at least once a yearto reelect board members and approve poli-cies. The boards have 7 normal membersplus alternate members. Board membersare elected for a renewable term of 2 years.Before transfer, there was a fixed quotasystem wherein 4 board members had tobe “small holders” whose total farm hold-ings in the district were less than 20 hect-ares each, and 3 members had to be “largeholders” whose total farm holdings in thedistrict were 20 hectares or more each.Both RUT and Rio Recio districts abol-ished this rule because farmers felt it wasarbitrary and exacerbated rather than less-ened tensions among WUA members. Keyinformants reported no severe social ten-sions among farmers of differing classes,landholding sizes, and wealth that wouldrestrict farmers to organize collectively.Corruption and misallocation of water oc-cur occasionally, but most farmers inter-viewed indicated that this was not a per-vasive problem and was related to a fewindividuals rather than classes.

As mentioned, the first transfers didnot convey full control over district bud-gets, O&M plans, water fees, and person-nel. However, these powers were finallyvested in the WUA after the implementa-tion of the Land Development Law of

10

TABLE 4.Powers devolved and functions transferred in sample districts.

Arrangements and functions RUT Rio Recio Samacá San Rafael Maria La Baja

Year of transfer 1990 1990 1992 1995 1996

WUA is legal entity Yes Yes Yes Yes Yes

WUA leaders selected by Yes Yes Yes Yes Yesfarmers

WUA has authority to make Yes Yes Yes Yes Yesrules and sanctions

Maximum sanction available to Fine, stop Fine, stop Fine, stop Fine, stop Fine, stopWUA service, take service, take service, take service, take service, take

user to court user to court user to court user to court user to court

Maximum sanction applied Take user to Stop service Fine Stop service Not yetsince transfer court

Authority to make O&M plan Shared with Shared with Shared with Yes Yesand budget* agency agency agency

Authority to set water charges* Shared with Shared with Shared with Yes Yesagency agency agency

Authority to hire or release Not initially Not initially Not initially Yes Yesmanagement staff*

Legal water right at level of No No No No Noscheme or farmer organization

Control over intake Yes Yes Yes Yes Yes

Control over main canal Yes Yes Yes Yes Not initiallysystem

Control over subsidiary canal Yes Yes Yes Yes Yessystem

Responsibility for future Not defined Not defined Not defined Not defined Not definedrehabilitation

Canal rights of way Yes Yes Yes Yes Yes

Right to make contracts and Yes Yes Yes Yes Yesraise additional revenue, not forprofit

WUA has right to make profits** No No No No No

* After 1993 requirement for agency oversight was discontinued.

** This issue is currently under consideration.

11

1993, which was enacted in 1994. Colom-bia has no distinct water law and thereare no water rights or concessions vestedin the districts or with individual farmers.Expectations about water entitlements toirrigation districts are based on precedentsabout maximum seasonal divertable dis-charges that have been specified by INATand previous irrigation administrations.In many cases, especially in water-abun-dant areas, or where there is only one di-version weir along a river, no specified ceil-ings exist. When a district is transferred toa WUA, it takes over full control of the ir-rigation network, including the intake. TheWUA has the right to make contracts withthird parties and raise supplemental rev-enue apart from water charges. Althoughthe issue is currently under consideration,WUAs are still prohibited from makingprofits.

Managerial Changes Made byWUAs after Transfer

At the heart of the theory of management devo-lution is the argument that, local users of aresource who are empowered as a groupto take over management of the resourcehave the incentive to manage the resourcemore efficiently and sustainably than acentrally financed government agency. Thevital element between adoption of transferand realization of impacts is the matter ofchanges in management by the WUA:What actual managerial changes, if any,are made by the WUA after transfer? Whatare the primary motivations for thechanges—to reduce costs, improve effi-ciency or reliability of management, or toenhance the productivity or profitability ofirrigated agriculture?

Table 5 identifies changes in manage-ment practices by WUAs in RUT, Rio

Recio, and Samacá districts following man-agement transfer.7 Enhancement of effi-ciency (especially cost reduction) was thedominant motivation for changes intro-duced in the three districts. This concernprompted such actions as streamlining or-ganizational structure, merging or abolish-ing positions, reducing staff, increasingwork hours and service areas, streamliningprocedures for fee payment, and introduc-ing computerized management informa-tion systems.

Table 6 displays information ondistrict staff deployment before and aftertransfer in the five districts. Three yearsafter transfer, the WUA managementreduced the number of staff in RUT, RioRecio, and Samacá by 20–60 percent.Post-transfer service area per staff rangedfrom 124 hectares for RUT (a 2-stage liftsystem) to 298 hectares in Samacá.

Other significant actions taken by theWUAs included replacing old “inherited”ditch tenders with new ones (who wereexpected to be more accountable to thenew WUA management). Supervision ofoperations was decentralized into zones.The districts began making structural re-pairs at their own expense (reportedly inmore pragmatic and cost-efficient waysthan had been done by public agencies be-fore transfer). The districts also began todiversify their revenue sources and hiredlawyers to collect overdue fee payments.These actions were intended to improveaccountability and competence of staff,management efficiency, financial solvencyof the district, and reliability of water de-livery. There is some question about thewillingness of farmers to invest in thelong-term sustainability of the irrigationschemes in that, after transfer, no districthas yet initiated an infrastructure replace-ment and improvement fund.8 Farmers in-dicated some unease about an accumula-

7San Rafael and MariaLa Baja are not includ-ed here because at thetime of this study notenough time had tran-spired after transfer formanagerial changes tobe realized.8Among districts trans-ferred in the 1990s, RUTwas the only one that es-tablished an equipmentreplacement fund. Coel-lo and Saldaña, trans-ferred in 1976, also setup such funds aftertransfer. However, nodistricts in the countryhave set up a capital-re-placement fund for ca-nals and water-controlstructures.

12

TABLE 5.Changes in management made by WUAs after transfer.

Type of change RUT Rio Recio Samacá Primary motivation

Streamlined organizational structure Yes Yes Yes Efficiency

General manager and board president Yes Yes Efficiencypositions merged

Small/large farmer board quotas eliminated Yes Yes Reduce tensions

Replaced ditch tenders with new recruits Yes Accountability

Reduced total number of staff Yes Yes Yes Efficiency

Staff work hours increased and made Yes Yes Yes Efficiencyflexible

Obtained limited training/consultation from Yes Yes Yes Staff competenceagency

Obtained limited training/consultation from Yes Yes Yes Staff competencethird party

Hired lawyer to recover overdue water fee Yes Yes Yes Financial solvency,payments accountability

Streamlined fee payment arrangements Yes Yes Yes Convenience, efficiency

Eliminated volumetric fee Yes Efficiency

Diversified revenue sources Yes Yes Financial solvency

Established capital replacement fund for Yes Financial solvencyheavy equipment

Sponsored improvements in canal network Yes Yes Reliability of waterdelivery

Established water management zones in Yes Accountability,scheme efficiency

Introduced MIS programs for administration Yes Yes Yes Efficiency,and finance accountability

TABLE 6.Deployment of district management staff.

Total district staff Service area per staff

District Before transfer After transfer Before transfer After transfer

RUT 92 78 105.4 124.3

Rio Recio 114 48 88.6 210.4

Samacá 13 10 229.3 298.1

San Rafael 5 5 118.0 118.0

Maria La Baja 94 92 98.5 100.6

Note: ‘After transfer data' for RUT, Rio Recio, and Samacá were taken 3 years after transfer. ‘After transfer data'for San Rafael were taken 19 months after transfer, and for Maria La Baja only 8 months after transfer.

13

tion of a large amount of money in a long-term account and many expected the gov-ernment to finance most, if not all, of thefuture costs of rehabilitation.

As indicated in figure 2 (charts 2.5and 2.6), except in Maria La Baja, the ma-jority of farmers interviewed felt that over-all administration of the district was goodboth before and after transfer. In RUT,Samacá, and San Rafael, 25 to 35 percentfelt it had improved after transfer. More

than half the farmers in Maria La Baja feltdistrict administration was poor beforeand after transfer. In RUT, farmers wereevenly split on whether communicationbetween farmers and district staff was ef-fective or not. In the other districts, mostfarmers reported a negative perceptionabout farmer/district staff communicationboth before and after transfer, or that it hadworsened (e.g., Rio Recio).

9In Samacá, the govern-ment began expenditurereduction before the1992 transfer, as part ofthe national policyadopted in 1990 to re-duce expenditures andtransfer management.10This includes the costof a hydroelectric ener-gy station in addition toirrigation system costs.It was not possible tosplit up the two costs inthis study.

Impacts of Transfer

Financial Impacts

The primary interest of the government forpromoting management transfer was todecrease government expenditures for therecurring costs of irrigation. In the shortrun, management transfer has reducedgovernment expenditures in the sampledistricts dramatically. As indicated in fig-ure 3, government expenditures per hectarefell from a range of US$20–US$80 per hect-are before transfer to zero or near zero af-ter transfer in the RUT, Rio Recio, andSamacá districts.9 During the same period(1985–95) government expenditures in SanRafael (the lift scheme that was onlytransferred in 1995) rose during itsdevelopment stage all the way to US$200–US$250 per hectare and have remained atthis high level to date.10 While expendi-tures for routine management costs havebeen eliminated or reduced substantially, itis not yet clear how much money thegovernment might provide to the districtsfor rehabilitation in the future. If aftertransfer WUAs begin to defer maintenancein anticipation of future governmentassistance for rehabilitation, it is possiblethat government expenditures for rehabili-tation could increase over time.

Transfer resulted in variable effects onthe cost of irrigation to farmers. In RUT,where the total cost of irrigation wasrelatively high (due to two-stage pump-ing), at transfer, farmers exerted pressureon their new board to contain costs. Over6 years after transfer, total farmer pay-ments declined in real terms by 22percent, from US$83 to US$65 per hectareby 1995. This was due to both a decline infee-collection rate, from above 90 percentbefore transfer to less than 70 percent by1995 (figure 4 and table 7) and a reductionin the amount of water delivered per hect-are (figure 5). Owing to the elimination ofthe government subsidy and decliningfarmer payments of water charges, theboard was pressured to increase the watercharge and reduce the budget in an effortto balance finances. The water charge wasincreased from US$67 per hectare in 1990(year of transfer) to US$108 in 1995(table 8) while the total O&M budget wascut by 42 percent between 1989 and 1995(table 7). However, these efforts were insuf-ficient due to underfinanced maintenanceand lobbying from farmers. By 1995, thegovernment again began paying for someof the cost of repair work.

14

FIGURE 2.Farmer perceptions about district performance before and after turnover.

60

2.1. Adequacy of water delivered.

0

20

40

Res

pons

es (

%)

Better afterWorse after

Both adequateBoth inadequate

2.2. Fairness of water distribution.

0

20

40

60

80

Res

pons

es (

%)

Better AfterWoser After

Both fairBoth unfair

0

20

40

60

Res

pons

es (

%)

Better afterWorse after

Both adequateBoth inadequate

2.3. Adequacy of maintenance.

0

10

20

30

40

50R

espo

nses

(%

)

Better afterWorse after

Both positiveBoth negative

2.4. Financial management.

0

20

40

60

Res

pons

es (

%)

Better afterWorse after

Both positiveBoth negative

2.5. Overall administration.

0

20

40

60

80

Res

pons

es (

%)

Better afterWorse after

Both positive Both negative

2.6. Effectiveness of communications.

Note: Both=Both before and after (e.g., Both adequate = Adequate, both before and after).

15

FIGURE 3.Cost of irrigation to the government and the farmers.

* Cost to the government = Funds spent, which originated from government sources other than the district.Cost to the farmer = Total water charges actually collected from farmers.

0

20

40

60

80

100

120

US

$/ha

Farmers Government

San Rafael

1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996

Tran

sfer

0

50

100

150

200

250

300

1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995

US

$/ha

Farmers Government

RUT

Tran

sfer

0

10

20

30

40

50

60

US

$/ha

Farmers Government

Rio Recio

1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995

Tran

sfer

US

$/ha

Farmers Government

RUT

0

20

40

60

80

100

120

140

160

180

Tran

sfer

1986 1987 1988 1989 1990 1991 1992 1993 1994 1995

16

FIGURE 4.Irrigation fee collection rates.

TABLE 7.Key performance indicators for sample districts, before and after management transfer.

RUT Rio Recio Samacá San Rafael

1989 1995 1989 1995 1989 1995 1989 1995

Relative water supply* ratio 2.1 1.9 2.3 2.1 1.4 1.6 1 1.1

Relative irrigation supply ratio 2 1.1 3.3 3.7 1.5 1.8 0.6 0.9

Water delivery capacity ratio 1.6 2.5 2.3 2.2 1.2 1.7 na na

Capacity utilization rate ratio 0.6 0.4 0.4 0.5 0.8 0.6 na na

Irrigation intensity % 154 153 142 114 106 149 197 190

GVO/Unit land* US$/ha 2,013 2,060 (1994) 2,155 2,112 (1994) 10,394 9,060 1,568 (1991) 1,808

GVO/Unit water* US$/m3 0.54 0.91 0.17 0.17 1.2 2.06 1.08 1.56

Fee collection rate % 82 67 na na 80 72 94 95

Financial self-sufficiency* % 51 69 19.4 114.9 29 107 21.2 52.8

Return on investment* % 25 33 31 29 12 22 24 28

O&M/Unit land US$/ha 163 95 38 54 111 61 180 256

O&M/Unit water US$/1,000m3 23 29 2 2 7 8 62 176

* These belong to IWMI’s standard minimum set of performance indicators (IIMI 1996).

17

FIGURE 5.Relative water supply (RWS) and relative irrigation supply (RIS) in sample districts.

1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995

Rainfall RWS RIS

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

1.60

RW

S &

RIS

0

200

400

600

800

1,000

1,200

mm

San Rafael

Tran

sfer

1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996

Rainfall RWS RIS

0.00

0.50

1.00

1.50

2.00

2.50

3.00

RW

S &

RIS

0

100

200

300

400

500

600

700

800

900

1,000

mm

Samacá

Tran

sfer

0.00

0.50

1.00

1.50

2.00

2.50

3.00

3.50

4.00

4.50

5.00

1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995

RW

S &

RIS

0

500

1,000

1,500

2,000

2,500

mm

Rainfall RWS RIS

Rio Recio

Tran

sfer

0.00

0.50

1.00

1.50

2.00

2.50

3.00

1986 1987 1988 1989 1990 1991 1992 1993 1994 1995

RW

S &

RIS

0

200

400

600

800

1,000

1,200

1,400

1,600

mm

Rainfall RWS RIS

RUT

Tran

sfer

18

TABLE 8.Water charges in sample districts before and after transfer.

District Area water charge Volumetric water charge Total water charge(US$/ha) (US$/1,000m3) (US$/ha)

1990 1995 1990 1995 1990 1995

Samacá 14.8 35.5 2.3 ** 23 36

Rio Recio 12.9 12.6 1.3 1.3 66 Pending

RUT 40.7 65.7 6.8 9.3 67 108

San Rafael 18.1 26.6 10.4 17.5 48 53

Maria La Baja 14.3 16.4 2.7 2.2 53 55

In Samacá, where total O&M costswere at the modest level of about US$20–US$25 per hectare, and where governmentsubsidies were eliminated completely, thecost of irrigation to farmers rose substan-tially after transfer. However, the newboard kept total O&M budgets at about thesame level after transfer. In Rio Recio, thedistrict was able to achieve financial self-sufficiency rapidly after transfer withoutraising its fees (which were nearly as highas in RUT, although it had no pumpingcosts). The district achieved this throughincreasing fee collections dramatically andavoiding increases in the O&M budget af-ter transfer. Although San Rafael was nottransferred until 1995, a trend of declininggovernment expenditures and rising coststo farmers occurred from the late 1980s tothe early 1990s. The district has not hadtrouble with fee collections from farmers,despite their extraordinarily high levels.This is largely because it is a pumpscheme and water is not delivered to indi-

vidual farms if payments by the respectivefarmer are in arrears.

The total cost of irrigation, or the totalannual O&M budget, did not increase sig-nificantly during the first few years aftertransfer in any of the schemes in the study.Figure 6 shows that the self-sufficiency offee levels (relative to O&M budgets) rose toapproximately 100 percent at transfer inRUT and Samacá, mainly due to drops inthe O&M budgets (which were followedby slight rising trends after transfer). Thiswas followed by worrisome declines in ir-rigation fee collection rates over severalyears after transfer (figure 4). By 1995, feescollected had dropped back down to only69 percent of the O&M budget (table 7). InRio Recio, the O&M budget remained atvirtually the same level for several yearsafter transfer, although the fee leveladequacy rose abruptly after transfer.This suggests that most of the additionalrevenues were not being appliedto O&M.

19

FIGURE 6.O&M budgets and financial self-sufficiency.

Rio Recio

1985 198619871988 19891990 19911992 199319941995 1996

O&M budget/ha Sufficiency of fees

0

50

100

150

200

250

300

US

$/ha

0

20

40

60

80

100

120

140

160

%

Tran

sfer

RUT

0

50

100

150

200

250

300

1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996

US

$/ha

O&M budget/ha Sufficiency of fees

Tran

sfer

San Rafael

0

50

100

150

200

250

300

1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996

US

$/ha

0

20

40

60

80

100

120

140

160

%

O&M budget/ha Sufficiency of fees

Tran

sfer

0

50

100

150

200

250

300

1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996

US

$/ha

0

20

40

60

80

100

120

140

160

%

O&M budget/ha Sufficiency of fees

Tran

sfer

Samacá

20

In the high-cost RUT district, the bud-get dropped abruptly in the year of trans-fer from about US$163 per hectare in 1989to US$95 per hectare in 1995. The abruptbudget cut was apparently not sustainable;as is also indicated by the just one-timerise in self-sufficiency of fees collected. Af-ter this, the budget again began to rise andself-sufficiency to fall. By 1995, only 69percent of RUT’s budget was mobilizedfrom fees, while the government again be-gan providing financial assistance to thedistrict.

Samacá had no trouble collectingenough fees to cover its total budget aftertransfer. In this sample, the gravity flowschemes were able to sustain financial sol-vency for the first 5 years after transfer,even where substantial increases in feeswere required. But neither of the high-costpump schemes (RUT and San Rafael) wasfinancially solvent either before or aftertransfer.

Figure 2 (chart 2.4) shows data onfarmer perceptions about financial man-agement of the districts before and aftertransfer. Except for Maria La Baja, themost common view among farmers wasthat financial management of the districtwas handled satisfactorily both before andafter transfer. However, a significant minor-ity of 10–25 percent had negative views. Inall cases, the majority perception was thattransfer has not brought about a signifi-cant change in the quality of financialmanagement of the districts. Data in figure4 indicate declining irrigation fee collectionrates in RUT and Samacá, which is an un-expected cause for concern. It suggests thatthe new farmer organizations lack ad-equate support from farmers.

Table 7 summarizes data on these fi-nancial and other performance indicators,before and after transfer. The financial self-sufficiency of fees collected to cover O&M

budgets increased dramatically in all cases(even San Rafael which was not turnedover), except RUT, which declined by al-most 40 percent. The decline in RUT is ap-parently attributable to its high cost andsudden loss of substantial subsidies fromthe government. Following transfer, expen-ditures for O&M in high-cost RUT andSamacá were cut substantially while, forlow-cost Rio Recio they increased signifi-cantly. O&M expenditures (by the govern-ment) in high-cost San Rafael increasedfurther even before it was transferred to theusers.

Quality of Operations

Monthly water balance data for thefour sample schemes are presented in fig-ure 7. The irrigation requirement (shownas hanging bars) is the difference betweeneffective rainfall and evapotranspiration.In all the four sample districts there aretwo marked rainfall periods during theyear. In RUT and Samacá, irrigation is re-quired to supplement effective rainfall innearly every month of the year (except No-vember in Samacá). In Rio Recio, irrigationis required only 8 months of the year. Noirrigation is required during the rainy pe-riods of April–May and October–Novem-ber. In all districts, irrigation is supplemen-tal to cover frequent and sometimeslengthy dry periods.

Figure 5 and table 7 present annualdata on Relative Water Supply (RWS),Relative Irrigation Supply (RIS), and totalprecipitation.11 RUT and Rio Recio haverelatively abundant water supplies in rela-tion to crop water requirements. Since RUTis a pump scheme, RWS and RIS valueswhich are above 2.0 suggest poor manage-ment in that they are pumping more thanwhat is required for irrigation. Prior to

11RWS is the sum of ir-rigation supply (IS) andtotal rainfall (TR) divid-ed by crop water re-quirement (CWR), i.e.,RWS = (IS + TR)/CWR.RIS is irrigation supply(IS) divided by the dif-ference between cropwater requirement(CWR) and effectiverainfall (ER), i.e., RIS =IS/(CWR-ER).

21

FIGURE 7.Average monthly water balance in sample districts.

Continued

22

FIGURE 7. (Continued)

23

transfer, farmers had got used to an on-de-mand system of ordering water and thiscontinued after transfer, despite the highcosts involved and the inefficiencies of thisdelivery system. However, after transfer,RIS has been on a declining trend as thenew management has attempted to reducethe amount of water pumped.

In Rio Recio, RWS and RIS valuesabove 3.0 also indicate that the manage-ment is diverting water far above crop wa-ter requirements, leading to greater ineffi-ciencies. However, since it is a run-of-the-river scheme and there are no other usersdownstream, there is little incentive for thedistrict to divert less water. But, given thatRio Recio has the opportunity to sell wa-ter for energy generation, the high RWSand RIS values are indications of high op-portunity costs. RWS and RIS have contin-ued to rise even after transfer.

From just before transfer in 1992 until1996, RWS and RIS values for Samacáhave been roughly within what can beconsidered an ideal level of approximately1.5. This did not change after transfer. Thelow values for RWS and RIS for San Rafaelindicate that farmers have been unwillingto pump unless it was absolutely neces-sary, given the high costs. Also, the pre-dominant crop is pasture, which is moreresistant to dry periods than crops grownin the other districts. However, values lessthan one indicate that the crops are beinggrown under stress and yields could beaffected. While it makes good sense ofmanagement to keep pumping costs down,this should not be done at the cost of cropyields.

No significant changes in RWS or RISwere observed in Samacá or San Rafael af-ter transfer. The significant decline in RISin RUT after transfer is explained mainlyby the high cost of pumping and the de-clining financial situation due to elimina-

tion of government subsidies and declin-ing revenues from water charges. Increasein RWS in Rio Recio after transfer is ap-parently more related to above-average lev-els of rainfall after transfer than to thetransfer itself.

Another measure of operational perfor-mance is the Capacity Utilization Ratio(CUR), expressed as a ratio of the annualpeak demand for irrigation to the maxi-mum canal carrying capacity. It is an indi-cator of the extent to which canal deliverycapacity is being availed. Data for RUT(shown in figure 8 and table 7) show agradual improvement from 1986 through1993, at which time the CUR reached1.25.12 The rise was related to shifts tocrops with higher water requirements. Thedecline after 1993 was apparently due tothe decline in RIS in recent years, broughtabout the financial crisis mentionedabove. Samacá also had a rising trend (fig-ure 8) over time that reversed at transferand stabilized at a moderate level of about0.6.13

Figure 9 shows the data on trends inirrigation intensity in the sample districts.San Rafael has the consistently highest in-tensity due to its production of pasturethat has grown through both cultivationseasons. Intensities increased to the rangeof 150–170 percent in both RUT (mainlybecause of increasing the area plantedwith sugarcane) and Samacá (mainly dueto increasing the area planted with mul-tiple crops of onion). But in both cases, theincreases began before transfer and can-not, therefore be attributed primarily tomanagement changes brought about bytransfer. Rio Recio has had a steady de-cline in irrigation intensity over the entireperiod, from 160 percent to 115 percent.This is related to a general shift fromflooded rice to cotton, sorghum, and othercrops. The decline in irrigation intensity

12Peak demand exceed-ed maximum capacityin 1993 (with a CUR of1.25). This happenswhen farmers establisha cropping pattern overan area such that the ir-rigation requirement ex-ceeds what can ade-quately be served withthe available canal ca-pacity. In this situationrotational deliveriescould be arranged and/or the delivered flowmay temporarily en-croach on canal free-board.13This was also appar-ently related to shifts incrop choices as noted inthe section entitled “Im-pacts on Productivity”p28.

24

despite high RWS and RIS is an indicationof poor water management, which hascontinued to worsen after transfer. This

statement is supported by the trend in val-ues of both water- and agriculture-relatedindicators, for this system.

FIGURE 8.Capacity utilization ratio in sample districts.

25

FIGURE 9.Annual irrigation intensity in sample districts.

14 A functional structureis defined as one thatcan currently performits basic design functionand shows no signs oflosing this capacitywithin about a year. Anearly dysfunctionalstructure is one that isconsidered to be likelyto become unable to per-form its basic functionwithin about one year’stime. A dysfunctionalstructure is one that wasunable to perform itsbasic function at thetime of the inspection.For canal reaches, dys-functional means it isunable to convey at least70 percent of the desiredflow capacity. Nearlydysfunctional means itis likely to become dys-functional within aboutone year’s time.

Figure 2 (charts 2.1 and 2.2) showsthat the majority of farmers interviewed inRUT, Rio Recio, and San Rafael felt thatwater delivered was both adequate andfairly distributed before and after transfer.Farmers in Samacá are divided betweenthe positive and negative views of wateradequacy and fairness of distribution. Thisis mainly due to the bifurcated and elon-gated structure of the system between thehillsides and valley floor and water-scarceconditions in Samacá.

We hypothesized that the proportionof farmers who perceive change in perfor-mance before and after IMT (or the refer-ence year for nontransferred schemes)would be different between transferredand nontransferred districts. This wasdisconfirmed at the 95 percent level ofprobability with the Pearson Chi-squaretest (X2), for the three indicators of:

1. Farmer perceptions about adequacy ofwater delivered to the farm(X2=0.0003),

2. Adequacy of system maintenance(X2=3.65), and

3. Fairness of water distribution(X2=3.76).

Our conclusion is that managementtransfer by itself did not bring about anyclear and significant changes in the qual-ity of irrigation operations in the studyschemes.

Impacts on Sustainability of Infrastructure

A detailed field inspection was made ofcanal lengths and structures in the RUT,Rio Recio, Samacá, San Rafael, and MariaLa Baja schemes. The study team in-spected sample canal lengths randomlyselected within upper, lower, and tailstrata along main, branch, and distributarycanals. All control and measurementstructures along the sample lengths wereinspected. Canal reaches and structureswere classified as ‘functional’ (F), ‘nearlydysfunctional’ (ND), and ‘dysfunctional’(D).14 Canal lengths were considered ‘de-fective’ if one of the following problems ex-isted and if it interfered with the desiredhydraulic operation—

Note: T-5 is 5 years before transfer. T is the year of transfer. T+5 is 5 years after transfer.

26

• constriction or enlargement of the ca-nal cross section

• visible siltation and/or encroachmentof freeboard or adjacent road

• visible seepage

• slippage, scouring, or other defects inthe embankment

• cracks or other damage to canal lining

Table 9 displays the percentage of to-tal canal length inspected in each schemethat was observed to be dysfunctional,nearly dysfunctional, and functional. Ex-cept for Maria La Baja (which had numer-ous design and construction problems andwas undergoing rehabilitation at the timeof the study), dysfunctional canal lengthswere rare. Nearly dysfunctional sectionswere found along 10 percent to 19 percentof observed canal lengths.

TABLE 9.Functional condition of structures inspected.

District Total structures Structures Dysfnctional Nearly dys- Functionalin scheme inspected (%) (%) functional (%) (%)

RUT 80 50 4 14 82

Rio Recio 234 17 0 3 97

Samacá 525 60 11 28 62

San Rafael* 40 25 0 12 88

Maria La Baja** 250 22 52 30 18

* Network inspected was drainage network. San Rafael has no surface irrigation canals.

** Some canals inspected are already under rehabilitation as part of the IMT process.

Control, conveyance, measurement,and ancillary structures were considereddefective (i.e., dysfunctional or nearly dys-functional), if one of the following condi-tions was present:·

• Scouring of canal adjacent to thestructure.

• The approach section, rubble pack, orwings of structures are breaking apart.

• The water control structure cannotcontrol flow as intended (due to gatesor sills missing, eroded, or damaged;significant leakage at gates; or damagedmechanism of movable structures).

• The water measurement structurecannot be used to measure flow due toa damaged or missing gauge, recorder,or other component.

• The civil works of ancillary structuresare damaged or poorly constructed.

Table 10 shows the number of struc-tures inspected that were found to be dys-functional, nearly dysfunctional, and func-tional. Again, Maria La Baja was in poorcondition with 82 percent of observedstructures being either dysfunctional ornearly so. This clarifies why the farmershave resisted transfer and insisted that thescheme be rehabilitated before transfer.Samacá is problematic with only 62 per-cent of observed structures being fullyfunctional. Figure 10 shows that most ofthe dysfunctional structures in Samacáare adjustable gates (many are missing ordamaged) and division boxes (many withbroken sills or walls).

27

TABLE 10.Functional condition of canal lengths inspected.

District Total canal Length Length dys- Length nearly Length func-network length inspected (%) functional (%) dysfunctional (%) tional (%)

RUT 170.7 10 0 17 83

Rio Recio 135.8 12 0 10 90

Samacá 58.0 28 6 19 75

San Rafael* 30.0 15 3 12 85

Maria La Baja** 284.7 13 19 19 62

* Network inspected was drainage network. San Rafael has no surface irrigation canals.

** Some canals inspected are already under rehabilitation, as part of the IMT process.

The study team working together withmaintenance staff of the Irrigation Depart-ment also estimated the cost using localmaterials and labor to repair all canallengths and structures that were identifiedas dysfunctional or nearly dysfunctional.In table 11, the cost estimate for repairingall dysfunctional canal lengths and struc-tures is referred to as the accumulated essentialmaintenance requirement (row 2). The cost esti-mate for repairing all defective canallengths and structures is referred to as theaccumulated preventive maintenance requirement(row 3). The combination of the two istermed the total accumulated or deferred

maintenance requirement (row 4), which meansmaintenance problems that have been de-ferred from routine maintenance.

If we add routine and accumulatedmaintenance requirements together (row 5)and compare this to routine expenditure,we have a value which is the percentageby which routine expenditure would haveto increase to take care of all routine, es-sential, and preventive maintenance re-quirements within 3 years (row 6). Clearly,the lower the value, the greater theorganization’s ability to handle currentmaintenance requirements without exter-nal subsidies.

FIGURE 10.Distribution of dysfunctional structures by type, Samaca irrigation system.

28

TABLE 11.Maintenance investment capacity in transferred districts.

Maintenance expenditures and requirements RUT Rio Recio Samacá(1995 US$ per ha)

1 Current average annual maintenance expenditure (US$ per ha) 38.18 25.45 19.12

2 Accumulated essential maintenance requirement (US$ per ha) 0.46 0 3.01

3 Accumulated preventive maintenance requirement (US$ per ha) 0.91 1.01 3.42

4 Total accumulated or deferred maintenance 1.36 1.01 6.43requirement (US$ per ha)

5 MIC Index (item 1/item 4) 28.10 24.95 2.97

6 Total annual maintenance expenditure required to do routinemaintenance plus complete essential maintenance in 1 year and 38.95 27.97 23.27preventive maintenance in 3 years (US$ per ha)

7 Increase of item 6 over item 1 (%) 2.0 1.3 21.7

15However, in RUT theinspection did not in-clude the condition ofpump stations, whichcould represent a signif-icant additional cost re-quirement.16Since no inspectionswere conducted beforetransfer, it is impossibleto directly comparefunctional conditions ofinfrastructure beforeand after transfer.

To eliminate deferred maintenance re-quirements within 3 years RUT and RioRecio would only have to raise the averagemaintenance budget by 2 percent and 1.3percent, respectively. This is quite reason-able and we conclude that maintenanceappears to be sustainable in these schemesfrom the aspect of resource mobilization.15

However, in Samacá, the situation ismore problematic. The district would haveto increase its maintenance budget by 21.7percent to eliminate the accumulated main-tenance requirement within 3 years. Intheory, this could be achieved either by in-creasing the fee collection rate above itscurrent level of 70 percent, or by increas-ing the fee per hectare, or both. The districthad already increased its water charges by56 percent between 1990 and 1995. Evenwith an additional increase in the waterfee, the fee will still be significantly lowerthan in the other districts. However, thereis cause for concern about whether the dis-trict will be able to raise the budget by afurther 21.7 percent to eliminate the back-log since the fee collection rate has been ona declining trend after transfer—from 80percent in 1993 to 70 percent in 1996 (fig-ure 4). An increase in the fee collection rate

could lead to a further decline in the col-lection rate.

Figure 2 (chart 2.3) shows that mostfarmers in RUT and Rio Recio perceivedmaintenance to be adequate before and af-ter transfer. Positive farmer perceptions inRUT and Rio Recio, and more dividedperceptions in Samacá, correspond gener-ally with the results of the scheme inspec-tions. Farmers in Maria La Baja are verydissatisfied with canal maintenance forreasons mentioned above. With the excep-tion of Maria La Baja, and some concernabout Samacá, we can conclude that mostof the schemes transferred appear to bephysically sustainable, assuming moderateincreases in maintenance investment canbe made.16

Impacts on Productivity

Changes in management practices and fi-nancial status of irrigation districts can beconsidered as direct outcomes of manage-ment transfer. These, in turn, could be ex-pected to have effects on the quality of op-erations and maintenance. Changes in thequality of operations and maintenance

29

could, in turn, effect the agricultural andeconomic productivity of irrigationschemes.

We have seen that, with the apparentexception of RUT, transfer did not result insignificant changes in the quality of opera-tions and it appears doubtful that it hadmuch effect on maintenance, as levels ofmaintenance investment have not deterio-rated (according to farmer reports). Hence,except for RUT, we would not expect tofind significant changes in agricultural oreconomic productivity that could be attrib-uted to management transfer.

Figure 9 shows that only one scheme(RUT) had significant improvement incropping intensity at the time of transferand afterwards. The cropping intensityrose from 110 percent to 160–170 percentafter transfer. The increase is attributable toa shift to sugarcane production (which isin the ground year-round), perennialgrape, and tree crops. However, soybeanyields (one of the major crops) remainedconstant before and after transfer (figure11). Also in RUT, gross value of output(GVO) per unit of land did not change sig-nificantly after transfer (figure 12). Thepost-transfer improvement in value of out-put per unit of water in RUT is attributableto the decline in RIS (as referred to above)

and it appears that the decline in RIS isrelated to financial pressures induced bytransfer.

In Samacá, irrigation intensity and po-tato yields (figures 9 and 11) were on therise before transfer and reversed to a slightdownward trend afterwards. But this isapparently more related to temporary dis-turbances caused by significant shifts incrop choices among farmers, rather than toeffects of management transfer. Substantialdrops in GVO per hectare and per cubicmeter followed by improvements aftertransfer (figures 12 and 13) are apparentlyattributable to temporary water balancedisturbances related to shifts by farmers tohigher-value commercial crops such as po-tato and onion. After the volumetric feewas discarded in Samacá in 1991, morepasture and vegetable crops were grownand there was a slight increase in theamount of water diverted per hectare (fig-ure 5). With the shift to higher value cropswhich occurred just before managementtransfer, the GVO per unit of land and wa-ter reversed to a rising trend after transfer.The discarding of the volumetric fee canbe linked to management transfer (in that,it was the decision of the farmer assembly),but changes in crop patterns towardsmore onion and potato cannot be linked.

FIGURE 11.Yields of major crops in sample districts.

30

FIGURE 12.Annual gross value of output per unit of land.

FIGURE 13.Annual gross value of output per unit of water.

31

Changes in the crop pattern in theother districts do not appear to be relatedto management transfer. Irrigation intensi-ties have been on a long gradual decline inRio Recio since before transfer, but haveremained stable in San Rafael (which wasnot transferred during the same period).Between 1985 and 1995, main crop yields(rice) in Rio Recio, which had been trans-ferred, and San Rafael (pasture) and Mariala Baja (rice), which were not transferred,have all remained fairly stable (figure 11).

Figure 12 shows that GVO per unit ofland remained relatively stable after trans-fer in RUT, Rio Recio, and San Rafael.Samacá was the exception. Figure 13shows that the GVO per unit of waterincreased significantly in three of thetransferred schemes, including non- trans-ferred San Rafael. In Rio Recio, althoughrice has been gradually replaced by cottonand other grains after transfer, the amountof water diverted has not declined; hence,GVO per unit of water has not improved.

As mentioned above, the increase inGVO per unit of water in RUT after trans-fer is due to a shift to higher value crops

such as sugarcane and smaller water or-ders due to higher pumping costs. Mainlydue to financial pressures created by theloss of government subsidies in RUT, andreductions in volume of water deliveredper hectare after transfer, the district sig-nificantly reduced its RIS ratio as well asits GVO per unit of water pumped into themain canals. In San Rafael, high pumpingcosts similarly resulted in higher GVO perunit of water delivered, but this cannot beattributed to management transfer, since itdid not occur until 1995. So we concludethat, except for RUT, management transferitself did not have significant impacts, ei-ther positive or negative, on the agricul-tural or economic productivity of the studyschemes.

Data are available on cost of irrigatio-nas a percentage of gross value of outputfor RUT, Rio Recio, and Samacá. In RUT,the declining trend in this indicator is re-lated to a decline in the RIS (as notedabove) and a decline in water fee collec-tion rates after transfer. Cost of irrigationrelative to GVO rose gradually after trans-fer in Rio Recio and Samacá (figure 14).

FIGURE 14.Cost of irrigation as percentage of gross value of output.

Note: Based on per ha farmer costs and output, in 1995 US dollar equivalents.

32

The national irrigation management trans-fer program in Colombia adopted in 1990can be characterized as a significant butonly partial devolution of management towater users. The government maintainedconsiderable advisory influence over thedistricts for several years, exercising somecontrol over O&M plans and budgets, andresisting district attempts to release largenumbers of staff. After adoption of the1993 Land Development Law, this controlwas relaxed considerably and districtsgained almost complete control over O&M,financial management, and disposition of

staff. However, powers devolved do not in-clude a formal water right or ownership ofirrigation scheme infrastructure. Also, thegovernment has not made it clear as towhose responsibility it will be, and underwhat terms and conditions, to finance pos-sible future costs of rehabilitation.

The findings of this study support thehypothesis that management transfer leadsto efforts by the WUA to improve manage-ment efficiency, such as reducing the num-ber of management staff and taking mea-sures to cut costs (see table 12 for a sum-mary of impacts).

Conclusion

TABLE 12.Summary of impacts.

Measures to improve cost efficiency High priority after transfer, several measures adopted

Measures to improve administration Majority of farmers report no changeand communications

Cost of irrigation to the government Significant reduction

Cost of irrigation to the farmers Upward trend, especially for pump schemes

Total cost of irrigation Mixed results, no clear pattern

Local financial self-sufficiency for O&M budget Increasing trend after transfer

Water distribution Mixed results. Planned improvements in relative irrigation supplyin 2 schemes. Majority of farmers report no changes inadequacy or fairness of water distribution

Financial sustainability of infrastructure Three sample schemes OK. Two others have significant deferredmaintenance or construction problems. Costly pump schemesare in doubt. No capital replacement funds set up

Functional condition of infrastructure Good in 3 schemes, problematic in 2 others

Agricultural productivity No significant changes

Economic productivity Mostly no change in GVO/ha. Improving trend in GVO/m3 water in 3schemes (partly related to managerial improvements in RIS in 2 schemes)

Ratio of cost of irrigation to gross value of Mixed resultsagricultural output

33

In the short term, the government hasachieved its objectives of significant reduc-tion of government expenditures for irriga-tion management. In most cases, this hasresulted in a significant but not severe in-crease to farmers in the cost of irrigation.In most cases, total costs of O&M did notchange much after transfer. The generalconclusion regarding operations and main-tenance is that management transfer didnot appear to bring about significantchanges in the quality of operations ormaintenance, that the transfer did not settleproblems, and that it did not interfere withfavorable management. Results of themaintenance survey suggest that mostgravity flow, river diversion schemes thatwere transferred in the early wave of thereform appear to have the capacity to fi-nance and carry out a sustainable mainte-nance program.

Schemes transferred more recently tendto have high management costs (such asSan Rafael) or dilapidated infrastructurewith high costs for repair (such as MariaLa Baja). This makes the prospect of localsustainability for these schemes muchmore problematic than those transferred inthe early wave of transfers. The financialsustainability of pump schemes is in doubtafter the government abolished all energysubsidies and farmers were forced to paythe full cost of pumping water. Of the fivedistricts, only RUT established an equip-ment-replacement fund. No district has setup a capital-replacement fund for basic in-frastructure. So there is cause for concernabout the willingness of the farmer-gov-erned districts to invest in the long-termsustainability of scheme infrastructure.

The sample districts did not show anysignificant improvements in agricultural oreconomic productivity that could be attrib-uted to management transfer. In summary,management transfer prompted a number

of managerial changes aimed at improvingmanagement efficiency and staff account-ability in the districts. Transfer resulted ina significant shift in the burden of costfrom the government to farmers, which hasgenerally been accepted by farmers. Buttransfer has not had substantial impactson the performance of operations andmaintenance, or on the agricultural andeconomic productivity of irrigated land orwater—neither improving negative perfor-mance nor causing detriment where per-formance is positive.

The assumption is commonly madethat management devolution will lead toimproved performance of irrigation sys-tems. But irrigation management transferis not a singular concept. There are a va-riety of strategies worldwide that vary indegree of authority transferred. In the caseof Colombia, management transfer wassignificant but not complete. In the earlyyears, the government maintained consid-erable authority over O&M plans, budgets,and district staff. The government retainsownership of infrastructure and is also re-sponsible for financing future rehabilita-tion and modernization of the schemes.No water rights have been granted to thedistricts or water users. Further compara-tive research is needed to test the hypoth-esis that a more integrated and compre-hensive devolution policy would lead tomore positive impacts on performance.

This study provides evidence that irri-gation management transfer does lead toimprove the efficiency and contain the costof management. This study suggests thatto be viable, local management should begiven full authority over O&M plans andbudgets, financial management, disposi-tion of staff, and enforcement of sanctions.The study lends credence to the argumentthat consistent and integrated policy sup-port is needed to induce farmers to invest

34

in the long-term financial and physicalsustainability of irrigation systems. Ex-amples of such kinds of support, whichare often lacking in management trans-fer programs in many countries, are thefollowing:

• Clear policy about terms and condi-tions whereby future subsidy or fi-nancial assistance for rehabilitation,or modernization of irrigation sys-tems will be provided.

• Linkage of such a policy to some de-gree of matching local investment,development of a long-term capitalreplacement fund by local manage-ment, and implementation of main-tenance according to agreed techni-cal standards.

• Arrangements whereby the govern-ment conducts technical audits ofmaintenance and helps guaranteethe financial integrity of capital re-placement funds.

• Involvement of farmer equity anddecision making in pre-transfer re-pairs to engender a sense of localownership in the irrigation system.

• Adequate on-the-job training pro-vided to new staff through at leastone or two full seasons prior totransfer.

• Special arrangements to provide ex-tra, but well-specified subsidies(which phase out gradually) to liftschemes which have high energycosts.

• A program to develop a post-trans-fer arrangement for provision of

support services to locally managedirrigation systems for technical and fi-nancial consultation, dispute resolu-tion, and political lobbying. Experi-ence in Colombia and elsewhere sug-gests that such services are needed af-ter transfer to help local managementbe viable in the long run.

Regarding this last point, in order tominimize costs and ensure a demand-driven approach, it may be worthwhile forgovernments to facilitate development ofprovincial or national-level water user net-work organizations, such as the Federriegoin Colombia (García-Betancourt 1995), toprovide such services after devolution.

Despite the last point above, it ap-pears likely that governments will stillneed to play roles in technical auditing orconsultation, and dispute mediation, pro-viding partial subsidies designed to stimu-late (not discourage) local investment, pro-viding financial security for long-termcapital-replacement funds raised byWUAs, water allocation, and water qualityregulation upstream. Perhaps the most fun-damental recommendation that could bemade at this point is that governments thathave adopted devolution policies but arestill designing programs to implementthem should develop integrated, compre-hensive devolution strategies that are in-ternally consistent and that promote theself-reliance of local management. Ad-equate investments should be made inmonitoring and evaluation of managementtransfer and, where necessary in action re-search, so as to ensure that the changeprocess is also an effective learning pro-cess.

35

ANNEX 1

Map of RUT District

RUTThe Roldanillo-La Union-Toro (RUT) irri-gation district is located in the prosperousCauca valley and serves 9,700 hectares. Itwas built between 1958 and 1971. Water ispumped from the Cauca river throughthree pumping stations for both irrigationand drainage. The district has predomi-nantly smallholdings, with 75 percent ofholdings being less than 5 hectares each.The main crops are cotton, grape, fruittrees, and sugarcane. As can be seen in fig-ure 7, rainfall is bimodal. The irrigation re-

quirement alternates with rainfall but isspread throughout the year. Water is deliv-ered on demand and is pumped a secondtime from the canal stet to farmers’ fields.Hence, pumping constitutes a major costto farmers.

Rio Recio

The Rio Recio district is located in theTolima valley in central Colombia, whichis one of the most productive agriculturalareas of the country. Constructed in 1951,it is a river diversion gravity flow scheme

ANNEX 2

Background Information on the Five Sample Districts

36

with open canals serving 10,200 hectaresof steep to undulating terrain. Averagelandholding sizes are the largest of the fiveschemes, with 75 percent of landholdingsbeing more than 10 hectares each. Maincrops are rice, sorghum and cotton, whichare irrigated on demand during periods ofwater abundance or on rotation duringwater shortages. The demand-based deliv-ery system, 234 control structures, andwater measurement structures down tofarm inlets constitute a highly flexible andmanagement intensive system.

SamacáThe Samacá irrigation district is located inBoyaca state, northeast of Bogota, in a nar-row valley rimmed with steep hills. It is atan elevation between 2,600 and 3,000meters. Water is supplied by two reservoirswith a combined total storage capacity of6.2 million cubic meters. Water is deliveredalong hillsides on both sides of the valleyand onto the valley floor. Irrigation is usedalmost year-round and is supplementalduring the two rainy periods of the year, inMarch–May and October–November (fig-ure 7). The district was constructed in1941 and presently serves 3,000 hectares,of which 46 percent is on hillsides and 54percent on the valley floor. Of all farmholdings in the scheme 95 percent is lessthan 5 hectares each. The main crops areonion, potato, pea, pasture, and veg-etables. Water delivery to farms is sched-uled weekly on the basis of written re-quests from farmers submitted to ditch ten-ders. On the hillsides, water is delivered tosmall temporary storage tanks, each serv-ing 10–20 hectares, from which water isdirected to individual fields through bur-ied plastic pipes that use gravity pressure.In the valley, water is normally directed tosmall temporary storage ponds at farm in-

takes, out of which water is pumped ontoeach field. Significant differences in the so-cioeconomic status exist between poorersmallholders on the hillsides and morewealthy and influential large holders inthe valley. However, regarding water, thesetensions have been partly offset by the dis-tribution of small farms upstream and aconsiderable amount of return flow fromthe hillsides being reused on larger farmsin the valley and at the tail end of the sys-tem.

San Rafael

San Rafael district is a small river liftscheme built in 1970. It irrigates 560 hect-ares of land located in the mountainousBoyaca State. Water is pumped from theChicamocha river (through a pump with acapacity of 0.60 m3/s) and delivered to in-dividual farms via a buried pipe system.Pasture and vegetables are the main cropsgrown. Soils and water have a high salin-ity content and much of the area is poorlydrained during rainy times of the year. Ir-rigation is virtually unnecessary duringthe peak rainy periods in April–May andOctober–November (figure 7). On the basisof farmer requests, which must be submit-ted at least 2 days in advance, district staffmake water delivery schedules. Before wa-ter can be delivered to a farm the area feeshould be paid up to date. Recently, thedistrict has been incorporated into thelarger Chicamocha irrigation district that iscurrently under development. San Rafaelis the first, and still the only unit in thelarger district to be functional.

Maria La Baja

The Maria La Baja irrigation district is lo-cated 50 km south of Cartagena City nearthe Atlantic coast. Its water supply comesfrom two reservoirs fed by two rivers.

37

Together, the reservoirs have a total storagecapacity of 235.5 million cubic metersand a combined discharge capacity of 20m3/s. The original design area was pro-jected to be 19,600 hectares, but 32 yearsafter construction in 1965, the actual irri-gated area is only 9,260 hectares. Siltationin the reservoirs, fed by deforestation inthe watershed, is cause for rising concern.The irrigated area has been shrinking inrecent years. The shortfall in expectationsand shrinkage has mainly been due to de-sign errors, flooding, poor drainage, andwaterlogging. There are also socioeco-nomic constraints such as extreme poverty.Bolivar State is in one of the poorer regionsof the country and small farmers are oftenundercapitalized. Seventy-five percent oflandholdings in the irrigated area is lessthan 10 hectares each. Sometimes, farmerscannot afford to purchase seeds or otherinputs or obtain credit. Also, a consider-

able amount of the potentially irrigableland in the scheme is used for cattle ranch-ing by wealthy farmers.

Topography in the irrigated arearanges from a slope of 12 percent near themountains to flat land in swampy, floodedareas at the tail end of the system. By theend of the rainy season, the water tabledepth is within 0.50 of a meter over 80percent of the area. On average, rainfall isrelatively high (1,890 mm per year) but itis highly irregular. The most importantcrop is rice. Sorghum, pasture, plantains,cassava, and maize are also grown. As inother districts in Colombia, water deliver-ies are based on requests received fromfarmers, with the precondition that fee pay-ments are up to date. It is reported thatthere are many problems with the designand construction of the scheme and manystructures are in disrepair.

38

Literature Cited

Alvarez, C., and C. Garcés-Restrepo. 1996. Perfil de Riego de la Republica de Colombia. Cali, Colombia: InstitutoInternacional del Manejo de la Irrigación.

Arriëns, Wouter Lincklaen, Jeremy Bird, Jeremy Berkoff, and Paul Mosley, eds. 1996. Towards effective waterpolicy in the Asian and Pacific Region. Volume 1— Overview of issues and recommendations. Proceedings ofthe Regional Consultation Workshop Towards a Policy for Water Resources Development and Manage-ment in the Asian and Pacific Region. Manila, Philippines: The Asian Development Bank.

Economic Development Institute of the World Bank (EDI). 1996. Handbook on participatory irrigation management.Washington D. C.: The World Bank.

Food and Agriculture Organization of the United Nations (FAO). 1994. Colombia: Agricultural sector review,irrigation and drainage studies. Rome: FAO and the FAO/World Bank Cooperative Programme.

Garcés-Restrepo, Carlos, and Douglas L. Vermillion. 1995. Irrigation management transfer in Colombia: As-sessment of seven transferred districts. In Irrigation management transfer, ed. S. J. Johnson, D. L. Vermillion,and J. A. Sagardoy. Water Report No. 5. Rome: Food and Agricultural Organization of the UN and theInternational Irrigation Management Institute.

García-Betancourt, Gilberto. 1995. The emergence of FEDERRIEGO: The Colombian federation of irrigation districts.Short Report Series on Locally Managed Irrigation. Report No. 8. Colombo, Sri Lanka: International Irriga-tion Management Institute.

Giraldo T., Alba Luz. 1997. Evaluacion de los Aspectos Sociales de la Tansferencia en Seis Distritos de Riego en Colombia.Project Report. Cali, Colombia: International Irrigation Management Institute.

IIMI. 1996. Quantification and measurement of a minimum set of indicators of the performance of irrigationsystems. Colombo, Sri Lanka: International Irrigation Management Institute. Duplicated.

Johnson III, Sam H, Douglas L. Vermillion, and Juan A. Sagardoy, eds. 1995. Irrigation management transfer.Water Report No. 5. Rome: Food and Agricultural Organization of the UN and the International IrrigationManagement Institute.

Ministry of Agriculture, Government of Colombia (MOA). 1990. El Desarrollo Agropecuario en Colombia. Bogotá,Colombia.

Plusquellec, Herve. 1989. Two irrigation systems in Colombia: Their performance and transfer of management to user’sassociations. Policy, Planning and Research Working Paper Series No. 264. Washington D.C.: The WorldBank.

Vermillion, Douglas L. 1997. Impacts of irrigation management transfer: A Review of the evidence. Research Report11. Colombo, Sri Lanka: International Irrigation Management Institute.

Vermillion, Douglas L., and Carlos Garcés-Restrepo. 1996. Results of management turnover in two irrigationdistricts in Colombia. Research Report No. 4. Colombo, Sri Lanka: International Irrigation ManagementInstitute.

Vermillion, Douglas L., M. Samad, N. Murthy, S. K. Raheja, U. Amarasinghe, and M. Svendsen. 1996. A standardmethodology to assess the impacts of irrigation management transfer. Draft. Colombo: International Irri-gation Management Institute. Duplicated.

World Bank. 1993. Water resources management. A World Bank policy paper. Washington, D.C.: The World Bank.

Research Reports

12. Water Distribution Rules and Water Distribution Performance: A Case Study in theTambraparani Irrigation System. Jeffrey D. Brewer, R. Sakthivadivel, and K.V. Raju,1997.

13. Rehabilitation Planning for Small Tanks in Cascades: A Methodology Based on RapidAssessment. R. Sakthivadivel, Nihal Fernando, and Jeffrey D. Brewer, 1997.

14. Water as an Economic Good: A Solution, or a Problem? C. J. Perry, D. Seckler, and MichaelRock, 1997.

15. Impact Assessment of Irrigation Management Transfer in the Alto Rio Lerma IrrigationDistrict, Mexico. Wim H. Kloezen, Carlos Garcés-Restrepo, and Sam H. Johnson III,1997.

16. Irrigation Management Transfer in Mexico: A Strategy to Achieve Irrigation DistrictSustainability. Sam Johnson III, 1997.

17. Design and Practice of Water Allocation Rules: Lessons from Warabandi in Pakistan's Punjab.D. J. Bandaragoda, 1998.

18. Impact Assessment of Rehabilitation Intervention in the Gal Oya Left Bank. Upali A.Amarasinghe, R. Sakthivadivel, and Hammond Murray-Rust, 1998.

19. World Water Demand and Supply, 1990 to 2025: Scenarios and Issues. David Seckler,Upali Amarasinghe, David Molden, Radhika de Silva, and Randolph Barker, 1998.

20. Indicators for Comparing Performance of Irrigated Agricultural Systems. David J. Molden,R. Sakthivadival, Christopher J. Perry, Charlotte de Fraiture, and Wim H. Kloezen,1998.

21. Need for Institutional Impact Assessment in Planning Irrigation System Modernization. D.J. Bandaragoda, 1998.

22. Assessing Irrigation Performance with Comparative Indicators: The Case of the Alto RioLerma Irrigation District, Mexico. Wim H. Klozen, and Carlos Garcés-Restrepo, 1998.

23. Performance of Two Transferred Modules in the Lagunera Region: Water Relations. G. Levine,A. Cruz, D. Garcia, C. Garcés-Restrepo, and S. Johnson III, 1998.

24. Farmer Response to Rationed and Uncertain Irrigation Supplies. C.J. Perry and S.G.Narayanamurthy, 1998.

25. Impacts of Colombia's Current Irrigation Management Transfer Program. Douglas L.Vermillion and Carlos Garcés-Restrepo, 1998.

Research Report

International Water Management Institute

INTERNATIONAL WATER MANAGEMENT INSTITUTEP O Box 2075, Colombo, Sri Lanka

Tel (94-1) 867404 • Fax (94-1) 866854 • E-mail iimi@cgiar.orgInternet Home Page http://www.cgiar.org/iimi

ISSN 1026-0862ISBN 92-9090-364-3

25

Impacts of Colombia’s CurrentIrrigation Management TransferProgram

Douglas L. VermillionandCarlos Garc s-Restrepoé