Regassa E. Namara, Bhawana Upadhyay and R. K. Nagar

IWMI is a Future Harvest Centersupported by the CGIAR

Adoption and Impacts ofMicroirrigation TechnologiesEmpirical Results from SelectedLocalities of Maharashtra andGujarat States of India

93

RESEARCHR E P O R T

SM

IWMI is a Future Harvest Centersupported by the CGIAR

Postal Address:P O Box 2075ColomboSri Lanka

Location:127, Sunil MawathaPelawattaBattaramullaSri Lanka

Tel:+94-11-2787404

Fax:+94-11-2786854

E-mail:iwmi@cgiar.org

Website:http://www.iwmi.org

I n t e r n a t i o n a lWater ManagementI n s t i t u t e

ISSN 1026-0862ISBN 92-9090-602-2

I n t e r n a t i o n a lWater ManagementI n s t i t u t e

Research Reports

IWMI’s mission is to improve water and land resources management for food,livelihoods and nature. In serving this mission, IWMI concentrates on the integrationof policies, technologies and management systems to achieve workable solutionsto real problems—practical, relevant results in the field of irrigation and water andland resources.

The publications in this series cover a wide range of subjects—from computermodeling to experience with water user associations—and vary in content fromdirectly applicable research to more basic studies, on which applied work ultimatelydepends. Some research reports are narrowly focused, analytical and detailedempirical studies; others are wide-ranging and synthetic overviews of genericproblems.

Although most of the reports are published by IWMI staff and their collaborators,we welcome contributions from others. Each report is reviewed internally by IWMI’sown staff and Fellows, and by external reviewers. The reports are published anddistributed both in hard copy and electronically (www.iwmi.org) and where possibleall data and analyses will be available as separate downloadable files. Reports maybe copied freely and cited with due acknowledgment.

i

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

Research Report 93

Adoption and Impacts of MicroirrigationTechnologies: Empirical Results fromSelected Localities of Maharashtra andGujarat States of India

Regassa E. Namara, Bhawana Upadhyay and R. K. Nagar

The authors: Regassa E. Namara is an Economist of the International Water ManagementInstitute (IWMI), Colombo, Sri Lanka (r.namara@cgiar.org) and Bhawana Upadhyay is anAssociate Expert of IWMI at its office in Anand, Gujarat, India (b.upadhayay@cgiar.org).R. K. Nagar is a Development Economist and Management Consultant in Agribusinessand Rural Development in Anand, Gujarat, India (rknagar@icenet.net).

Acknowledgments: This study was supported by the Comprehensive Assessment of WaterManagement in Agriculture and IWMI-TATA Water Policy Programs.

The authors appreciate the cooperation extended by farmers who participated in the survey.We particularly thank Maria Saleth and R. Sakthivadivel for their reviews and invaluablecomments, and David Molden, Tushaar Shah, Hugh Turral and Francois Molle for theircomments on the earlier drafts of the paper. The authors also thank the research anddevelopment staff members of Jain Irrigation Systems Ltd. for their information onmicroirrigation issues in Jalgaon and in India in general.

IWMI receives its principal funding from 58 governments, private foundations, andinternational and regional organizations known as the Consultative Group onInternational Agricultural Research (CGIAR). Support is also given by the Governmentsof Ghana, Pakistan, South Africa, Sri Lanka and Thailand.

Namara, R. E.; Upadhyay, B.; Nagar, R. K. 2005. Adoption and impacts of microirrigationtechnologies: Empirical results from selected localities of Maharashtra and Gujarat statesof India. Research Report 93. Colombo, Sri Lanka: International Water ManagementInstitute.

/ microirrigation / drip irrigation / sprinkler irrigation / irrigation systems / technology /models / economic analysis / yields / cropping systems / water conservation / food security/ poverty / women / India /

ISSN 1026-0862ISBN 92-9090-602-2

Copyright © 2005, by IWMI. All rights reserved.

Cover photo by Mats Lannerstad shows an Alfalfa field irrigated by a microirrigation systemin northern Gujarat, where dairying is the mainstay of the livelihood of the farming population,and a farmer irrigating his field using the conventional method.

Please send inquiries and comments to: iwmi@cgiar.org

iii

iii

Contents

Summary v

Introduction 1

Objectives 2

Background 3

Data and Analytical Methods 9

Analyses of Microirrigation Adoption 14

Agro-economic Analyses of Microirrigation Technologies 20

Other Impacts of Microirrigation Adoption 28

Conclusions and Implications 31

Literature Cited 35

Appendix 1 37

Appendix 2 41

Acronyms

AKRSP Aga Khan Rural Support ProgramAPPROTEC Appropriate Technologies for Enterprise CreationDAP Diammonium phosphateIDE International Development EnterprisesMPP Marginal Physical ProductNGO Nongovernmental organizationSD Standard DeviationVMP Value of Marginal Product

Glossary

Kharif July-October crop seasonpanchayat Institution of local self-governance in India, which is structured at district, block and village

levels. A panchayat and its representatives have a uniform five-year term.Rabi November-April crop seasonsummer season April/May-July crop seasontaluka A subdivision of a revenue district

v

Summary

Microirrigation technologies are aggressivelypromoted in India by the central government,state governments and many nongovernmentalorganizations (NGOs), both local and interna-tional, by providing different kinds of financial,institutional and technical support systems.These technologies are promoted primarily for oneor more of the following reasons: (1) as a meansto save water in irrigated agriculture, (2) as astrategy to increase income and reduce poverty,and (3) to enhance the food and nutritionalsecurity of rural households. Despite the reportedsignificant economic advantages and the con-certed support of the government and NGOs, thecurrent microirrigation area in India remains aninsignificant proportion of its potential. Based onthe data from recent field studies in Maharashtraand Gujarat, this report analyzes: (1) the econom-ics of alternative microirrigation technologiesranging from low-cost drip and sprinkler systemsto the capital-intensive systems, (2) the determi-nants of adoption of microirrigation technology, (3)the poverty outreach of the differentmicroirrigation systems, and (4) the sustainabilityimplications of microirrigation adoption. In linewith the findings of numerous other studies, thisstudy indicates that microirrigation technologies

result in a significant productivity improvementand, hence, economic gain over the traditionalmethod of surface irrigation. It also shows thatthe productivity gain of conventional drip systemsis significantly higher than that of low-cost dripsystems. Thus, low-cost microirrigation systemscannot be regarded as ends in themselves but asstepping stones for adopting the conventionalsystems, which are technically robust andeconomically more rewarding. The most importantdeterminants of microirrigation adoption includeaccess to groundwater, the prevailing croppingpattern, level of education, financial resources,the social stratum of the household, and thewealth or poverty status of the farmer. Contrary toexpectations, the majority of the current users oflow-cost microirrigation systems belong to thericher section of the farming population. Thestudy also indicates that the impact ofmicroirrigation systems on the long-termsustainability of groundwater resources dependson the magnitude of the overall productivity gainfollowing the shift from surface irrigation tomicroirrigation, the behavior of the adoptersfollowing the shift or the pattern of use of thesaved water, and the type and potential numberof adopters.

1

Adoption and Impacts of MicroirrigationTechnologies: Empirical Results from SelectedLocalities of Maharashtra and Gujarat States of India

Regassa E. Namara, Bhawana Upadhyay and R. K. Nagar

Introduction

In many parts of the world, the demand foravailable water resources is fast exceeding thesupply and competition between the varioussectors of the economy for scarce water isbecoming intense. In response to theseconditions, policymakers, researchers, NGOs andfarmers are increasingly pursuing variousinnovative, technical, institutional and policyinterventions to enable the efficient, equitable andsustainable utilization of scarce water resources.Microirrigation technologies constitute an elementof such innovative intervention approaches.Originally, microirrigation was often associatedwith the capital-intensive, commercial farms ofwealthier farmers. The systems used on largefarms, however, are unaffordable for smallholdersand are not available in sizes suitable for smallplots. Recently, these technologies have gonethrough technical transformations from largelycapital-intensive features to an input mode.

A survey of literature on the impacts ofmicroirrigation technologies indicate that they areusually promoted primarily for one or more of thefollowing objectives: (1) as a means of savingwater in irrigated agriculture and averting theimpending water crises (Narayanamoorthy 2003;Polak et al. 1997; Shah and Keller 2002), (2) as astrategy to increase income and reduce povertyamong the rural poor, (3) to enhance the food andnutritional security of rural households (Bilgi 1999;

Upadhyay 2003; Upadhyay 2004), and (4) as ameans to extend the limited available water overa larger cropped area, especially during droughtyears or during the period before a monsoonseason. Microirrigation technologies lead topoverty reduction through substantial increases infarm income due to an increased area ofcultivation, better crop yields, enhanced outputquality, early crop maturity and hence higher unitprices, and reduced cultivation costs, particularlyfor operations like irrigation and weeding.Microirrigation technologies enhance nutritionalsecurity by enabling the production andconsumption of vegetables, particularly leafyvegetables, which are usually missing in thetraditional staple diets of many cultures.

There are two lines of thought regarding thewater-saving potential of microirrigationtechnologies. The first line of argument is that theadoption of microirrigation technologies results innet water savings, thereby easing the prevailingwater-scarcity problems. The water saving isattained through substantial reduction in lossesdue to evaporation and inefficient fieldconveyance and distribution systems.1 Forinstance, water application can be reduced by 50to 100 percent through the drip method ofirrigation. This is the declared motive for the stategovernments of India to embark on the massivepopularization of these technologies. However, the

1For more information on this see Sivanappan 1994 quoted in Narayanmoorthy 1997.

2

farmers’ rationale for adopting these technologiesmay be different from the policy objectives of thestate governments. Farmers may give moreweight to the other attributes of microirrigationtechnologies such as improvements in yield,reduction in labor requirement, improvement inoutput quality, etc. in their adoption decisions.

The second line of thought is that eventhough microirrigation technologies can result inwater savings at the plot or field level, it may nottranslate into net water savings at a higher levelof aggregation such as the watershed or thebasin (Molden et al. 2003). According to this lineof thought, the net water savings could be onlymodest if the phenomenon of return flows, muchof which goes to recharge the underground watersource, is considered as useful. Thus theadoption of microirrigation technologies may notautomatically lead to water savings at the basinlevel unless enabling institutional and economicpolicy instruments are put in place that allow theequitable distribution or allocation of the savedwater.

Various studies in India have shown aconsiderable return to farmers’ investments inmicroirrigation technologies (Dhawan 2002;Narayanamoorthy 1997; Narayanamoorthy2003; Verma et al. 2004). For instance,Narayanamoorthy reported an increase in farmbusiness income of 64.6 percent, 79.5 percentand 83.4 percent for banana, grape andsugarcane, respectively for drip irrigationadopters in the Maharashtra state of India.Substantial efforts have been made todisseminate and popularize these technologies.For instance, state governments of India haveencouraged private involvement in themanufacturing and distribution of thetechnologies and adoption by farmers throughtargeted subsidy schemes. Despite theseefforts, however, the area under currentmicroirrigation systems remains an insignificantproportion of the potential. Thus, finding outwhy microirrigation technologies are notdisseminating fast and to the extent anticipatedis an important research issue.

Objectives

This report presents the results of a study doneto assess the adoption and impacts ofmicroirrigation technologies in selected villages ofMaharashtra and Gujarat states of India. Datawere collected in two phases. First, focus groupinterviews and key informant surveys wereundertaken. Second, relevant household, farmand field level data were gathered throughstructured questionnaire surveys.

The specific objectives of the study were to:

1. Analyze the economics of alternativemicroirrigation technologies ranging from low-

cost drip and sprinkler systems to capital-intensive systems.

2. Identify the determinants of microirrigationadoption and assess their quantitativeimpacts.

3. Evaluate the poverty outreach of the differentmicroirrigation technologies.

4. Discuss other impacts of microirrigationadoption.

3

Study Locations

The study was conducted in the Rajkot andJunagadh districts of the Gujarat state (figure 1)

Background

and in the Jalgaon district of the Maharashtrastate of India (figure 2). The salient features ofthe study locations, including the list of the studyvillages, are given in Appendix 1.

FIGURE 1.The study districts (and talukas) of Gujarat state.

4

FIGURE 2.The Jalgaon district of Maharahstra showing the five talukas in which the study villages are located.

In Jalgaon district, 14 villages in five talukaswere sampled from a total of 1,510 villagesspread over 13 talukas in the district (seeAppendix 1, table A1). The average annual rainfallin the district is 747.7 mm, of which 87 percentoccurs during the monsoon months of June-September. The highest rainfall occurs duringJuly. The total perennial river length in the districtis 730 km.

In Junagadh district, 6 villages from theMaliya taluka were selected for the study. In this

district, out of 1,034 inhabited villages, 966 areelectrified. Moreover, all the villages have accessto drinking water in wells or rivers or to pipe-bornewater. And public transport is available in 90percent of the villages. Similarly, in Rajkotdistrict, progress in rural electrification isremarkable with 854 out of 856 inhabited villageshaving access to electricity. All the villages haveaccess to drinking water in wells or canals or topipe-borne water. However, wells are the mainsource of drinking water. Public transport is

5

available in 97.4 percent of villages. All thevillages of the Kotada Sangani and Jetpur talukashave public transportation.

The proportion of the rural population is higherin Junagadh and Jalgaon districts. The populationdensity for the study districts varies from 200 to316 people per km2 and Jalgaon was the mostdensely populated district in 2001 with 316 peopleper km2 (see Appendix 1, table A2).

In all the study districts, the gross crop areahas increased over the years from 1992 to 2001.The percentage of cultivated area ranges from 47to 74 percent with Jalgaon recording the highestcultivated area. The cropping patterns of thestudy locations during the 1991/92 seasons aregiven in Appendix 1, table A3. There areimportant differences in the cropping patternamong the three districts and between the twostates. Oil crops, specifically groundnuts, are themost important crops in Rajkot and Junagadhdistricts, while in Jalgaon cereals followed bycotton are the most important crops. Fruit cropproduction (especially banana) is significant inJalgaon. The banana cultivated area in Jalgaonalone constitutes over half of the total bananaarea in Maharashtra state. Fruit crop production isrelatively insignificant in the other two studydistricts.

A comparison of cropping patterns of the twostates shows that cereals, pulses and fiber crops,in that order of significance, are the importantcrops in Maharashtra. Cereals, oil crops, foddercrops, and fiber crops, in that order ofsignificance, are the most important crops inGujarat. The significance of fodder crops inGujarat is an indication of the importance of thelivestock (dairy) sector in the farming system.Livestock occupies a pivotal place in the ruraleconomy of Junagadh and Rajkot districts, withcows and buffaloes making up the highestpercentage of livestock. Milk dairies andcooperatives thus contribute significantly to theeconomy of these regions. Moreover, the fishingindustry supports many households in the coastalareas of Junagadh.

The percentaged irrigated area for the studydistricts ranges from 16 to 26 percent and the

lowest irrigated area is observed in Jalgaon.When the two states are compared, thepercentage of irrigated land is higher in Gujaratthan in Maharashtra. Except in Rajkot, the totalarea under irrigation has shown an upward trendduring 1992 to 2001. In Rajkot, the total irrigatedarea has decreased slightly due to a sharpdecline in well irrigation. In fact the two studydistricts in Gujarat are counted among the fivedistricts of Gujarat that have recorded a declinein water level. During 1982-2001, all the studydistricts experienced a decrease in water level ofmore than 4 meters. Furthermore, during 1999-2002, 5 districts of Gujarat and 27 of Maharashtrarecorded a decline in water level. Along with theobserved decline in water level and abandonmentof some wells, there is well drilling activity. As ofNovember 2000, the number of wells drilled inGujarat and Maharashtra were 579 and 693,respectively.

Institutional Support Systems forMicroirrigation TechnologyDissemination

A number of NGOs, governmental organizationsand private business firms are engaged in thepromotion of microirrigation technologies in thestudy locations. The most prominent NGOsoperating in the region are the Aga Khan RuralSupport Program (AKRSP) in Gujarat andInternational Development Enterprises (IDE) inboth Maharashtra and Gujarat. IDE’s work isfocused mainly on designing or redesigningmicroirrigation technologies to make them morepoor-friendly and creating awareness amongfarmers about the new designs throughpromotional activities like meetings with farmers,video shows, field demonstrations, exhibitions invillage markets and so on. IDE does notsubsidize the financial investment for acquiringthe technology. They may, however, connect thefarmers to financial institutions.

A number of private firms are engaged inmanufacturing, marketing and distributingmicroirrigation systems. All in all, there are over

6

75 companies in India manufacturing and sellingdrip irrigation systems.2 The clients for thesefirms are primarily the highly commercializedlarge farmers, specializing in cash cropproduction. However, intense competition hasdriven new attempts to customize irrigationtechnology to suit the budget constraints of thesmall farmer in an effort to expand marketoutreach.

Among these firms, Jain Irrigation Systems iswell known among the farmers of Jalgaon. Thespecial feature about this firm is that it is ahomegrown company while the others aresubsidiaries of foreign companies, mainly fromIsrael. Jain Irrigation Systems providescomprehensive services, including engineering(manufacturing), assembling, marketing ordistribution, research and farmer training. Theresearch component includes soil analysis andproduction of tissue culture of banana rhizomes.In addition to irrigation systems, the companyalso provides input and output marketingservices. They distribute improved seeds orplanting materials, fertilizers and pesticides andbuy farm output from farmers. They are alsoengaged in the processing and export markets.Jain Irrigation Systems’ managers claim thatJalgaon is the most arid area in which bananacultivation has been made possible throughinnovation in microirrigation.

Private firms and NGOs have their ownelaborate marketing systems and channels. Inany case, the main participants in the supplychain include manufactures, dealers, distributors,

assemblers, extension volunteers and farmers.These market participants operate at differentspatial scales ranging from village to taluka ordistrict. The NGOs and government institutionsfacilitate market efficiency through provision ofinformation and subsidies. A typicalmicroirrigation-marketing channel in Gujarat whereAKRSP is involved is depicted in table1.

AKRSP recruits and trains Assemblers andVillage Extension Volunteers. The purpose of theVillage Extension Volunteer is the disseminationof information and identification of farmersinterested in trying the technology. TheAssembler prepares a proposal based on thefeedback from a Volunteer to send to AKRSP forapproval. The AKRSP approves the proposalbased on an assessment by AKRSP’s owntechnical staff. The Assembler then obtains theequipment and parts direct from the dealer andinstalls the system in the farmer’s field. Followingsuccessful installation, the subsidy is releaseddirect to the farmer. Although this is the normalchannel, in many cases, the farmer may dealeither directly with the Assembler or with AKRSP.

The subsidy given to farmers depends on thetype of the institutions involved in promotion.AKRSP has two subsidy schemes depending onthe quality of water used for agriculture. Forfarmers operating in saline areas, the subsidyrate is 50 percent of the cost while thecorresponding figure for sweet water areas is 33percent. The Government of Gujarat, in itsscheme for better use of water in irrigation, hasauthorized different subsidy rates depending on

TABLE 1.An example of a microirrigation technology supply chain in AKRSP-assisted areas.

Channel 1 AKRSP Farmer

Channel 2 Assembler Farmer

Channel 3 Assembler AKRSP Farmer

Channel 4 Dealer/Supplier Assembler Extension Volunteer Farmer

Note: AKRSP = Aga Khan Rural Support Program.

2Some of these are: Netafim Irrigation India Pvt. Ltd., Finolex group, Plastro Plasson Industries India Ltd., Plastro Irrigation India Ltd.,Balson Polyplast Pvt. Ltd., Naan Irrigation System, Pragati Microirrigation System, Jain Irrigation Systems, and Ganesh Irrigation.

7

the socioeconomic status of the farmer and themodel of microirrigation technology. For small,marginal, backward, tribal and female farmers, a50 percent subsidy is provided for procuring drip,sprinkler and pipeline systems. Large farmers andrecognized social and educational institutionslinked to agriculture get a 35 percent subsidy fordrip systems, 33 percent for sprinklers and 40percent for pipelines.

Since 2002, the Government of Gujarat hasset more restrictive preconditions regardingeligibility for subsidy incentives. According to anew directive, a subsidy is given for a continuousblock of 5 villages or a block of 100 hectares. Inother words, a group of 50 beneficiaries have tobe formed to apply for a subsidy. The intention ofthe new approach is to ensure that sufficientdemonstration of trust in the technology developsto encourage an early spread of the technology.However, this approach has changed the adoptionpattern from an individual decision-makingprocess to a collective one. Similarly, theGovernment of Maharashtra was earlier providinga 50 percent subsidy for microirrigation adoptionfor up to 2 hectares of land per farm. Thisscheme was withdrawn in 2002 but wasreintroduced at a 25 percent rate in October 2003.Following the withdrawal of the subsidy scheme,some of the microirrigation suppliers, such asJain Irrigation Systems, experienced a sharpdecline in sales.

In addition to the state government subsidyschemes, there are also central programs for thepromotion of microirrigation technologies. UnderIndia’s last Five-Year Plan, the subsidy was setat 25 percent for all categories of farmers, downfrom the previous 50 percent rate. The centralprograms are commodity-based. Among these,the central scheme for the use of plastics inagriculture (horticultural development) provides a25 percent subsidy for all forms of microirrigationtechnology. Likewise, the central subsidy schemefor the integrated development of vegetables,including root and tuber crops, provides anincentive of a 25 percent subsidy.

Many farmers consider the subsidy procedureas a very cumbersome undertaking. For example,

a farmer has to first approach a dealer and paythe full cost of the system to the dealer to getthe drip system installed. Only after installationcan a farmer make an application for a subsidy tothe district panchayat along with the requireddocuments like proof of installation and landownership. An officer from the panchayatrecommends and sends the proposal to a deputydirector only after inspecting the field to confirmthe installation of the system. Finally, the relevantdeputy director approves the subsidy and a checkis sent directly to the farmer. This procedure isseen as a complicated one by many farmers asthey have experienced difficulty in meeting theinitial expenses to buy the system. The farmerscomplain that the subsidy is not being provided atthe time when it is most needed. Thus thesignificance of the subsidy scheme to the cash-constrained, poor farmers is particularly limited.

Microirrigation Systems and theirTechnical and SocioeconomicAttributes

Different kinds of traditional surface irrigation andmicroirrigation systems are found in the studyvillages (table 2). Among the surface irrigationmethods, flooding is most common in the samplevillages of Gujarat while the furrow system ismore prevalent in the Maharashtra samples. Fewfields are under rain-fed farming. Thus, theproportion of area under irrigated farming in thesampled villages is way above the average levelsfor the study districts and the two states. Theproportion of rain-fed fields is higher inMaharashtra than in Gujarat, including fields ofcrops such as maize, sorghum, pulses and oilseeds. Microirrigation technologies can becategorized into two groups based on theirtechnical, economic and social attributes. Theseare low-cost microirrigation technologies and thecommercialized, state-of-the-art microirrigationsystems. The low-cost microirrigationtechnologies include the “pepsee,” easy drip,various kinds affordable microirrigation systemsdesigned by IDE, microsprinklers and microtube

8

TABLE 2.Types of irrigation systems observed in the study locations.

Irrigation system Maharashtra Gujarat

Area (ha) % Area (ha) %

Surface irrigation and rain-fed systems

Flooding 58.3 9.3 440.0 53.8

Furrow 129.5 20.6 80.1 9.8

Ring or round method 0.0 0.0 9.4 1.1

Rain-fed 66.4 10.6 6.0 0.7

Microirrigation systems

Pepsee or easy drip (AMIT) 16.6 2.6 27.0 3.4

Microsprinklers 0.0 0.0 90.2 11.0

Microtube drip 271.6 43.3 91.6 11.2

Conventional drip 85.4 13.6 30.8 3.7

Conventional sprinklers 0.0 0.0 43.5 5.3

Number of fields 327 461

Note: AMIT = IDE’s Affordable Microirrigation Technologies

drip systems. The latter class of microirrigationtechnology includes conventional drip andsprinkler systems.

The lack of adopters of conventional sprinklerand microsprinkler irrigation systems among oursample farmers from Maharashtra does not meanthat these technologies are totally absent in thestate. There are other villages in Maharahstrawhere sprinklers are in use.

The technical, economic and social attributesthat distinguish low-cost microirrigation systemsfrom commercial, state-of-the-art microirrigationsystems are as follows.

1. Affordability: The dominant feature of the low-cost microirrigation systems is that theyrequire little initial capital for successfuladoption, which has a high opportunity cost inthe context of the study locations. Thereduction in capital intensity was realizedthrough the replacement of conventionalemitters with microtubes, using a clean pieceof cloth in place of the filter, removing thenutrient-tank components and valves, andreplacing the pipe and lateral network ofconventional irrigation systems by a networkof low-cost tubes.

2. Local manufacturing capacity: The systemsare designed with the best availablecomponents, with preference given to localmanufacturing that only requires relativelyunsophisticated facilities, but not at theexpense of affordability and functionality.

3. Payback period: The income generationpotential of the systems (compared to thesystems they replace) at least covers theinvestment cost in one irrigation season.

4. Compatibility with the farming system: Thesystems are compatible with smallholderfarming systems because they are availablein a range of small packages from as little as20 square meters to a couple of hectares.They are also expandable, so the area servedcan be enlarged as farmers gain confidencein the technology and become morefinancially capable.

5. Pressure requirement: The required inletpressure head for the low-cost drip systemsranges from 1 to 4 meters, while theconventional systems require high pressure,careful filtration and special acid treatment tokeep it operating properly.

9

6. Ease of technical understanding by users:The systems are simple and easilyunderstood, and operated and maintained byaverage users, while the conventionalsystems are sophisticated and need technicalexpertise.

7. Uniformity of irrigation application: Lessuniformity as compared to the state-of-the-artmicroirrigation systems.

8. Operational convenience: The low-costmicroirrigation technologies have lowoperational convenience compared toconventional systems. For instance, installinglow-cost drip systems requires higher laborinput.

9. Compatibility with local micro-entrepreneurship: The skills required forassemblage and servicing of the systems arenot high and are compatible with those oflocal micro-enterprises. The design, serviceand maintenance costs are also low.

The successful adoption of microirrigationtechnologies requires the fulfillment of three basicfactors: (1) the technologies need to betechnically and economically efficient, (2) thetarget beneficiaries need to be aware of orknowledgeable about the technical and economicsuperiority of these technologies, and (3) thetechnologies must be accessible to potentialusers.

Data and Analytical Methods

The data for this study were gathered through asurvey of 448 farm households selected atrandom in Gujarat and Maharashtra, duringSeptember and October 2003. A structuredquestionnaire survey was preceded by focusgroup discussions and key informant surveys inthe study locations. The following analyticalapproaches were used to analyze and interpretthe resulting data.

Logit Adoption Model

To facilitate the choice of an appropriateeconometric model for analyzing the determinantsof microirrigation adoption and their quantitativesignificance, a precise definition of microirrigationadoption or adopter is required. In the presentcontext, microirrigation adopters are thosefarmers who use one or more of the

microirrigation technologies being promoted at thetime of the study by the government or NGOs onall or part of their fields. In this case, themicroirrigation adoption variable is a discrete-dichotomous variable (a farmer is either amicroirrigation adopter or a non-adopter). Thus thedefinition includes partial adopters. The non-adopters, or non-microirrigation farmers, are thosewho have not used microirrigation during the yearof the survey.

In instances where the adoption variable isbinary (0/1), logit and probit models are mostcommonly used to analyze technology adoptionprocesses (Aldrich and Nelson 1984; Feder et al.1985). Here the logit model is used to explain themicroirrigation adoption process.3 Thespecification of the logit model is as follows:

3For an explanation of the differences and similarities between these two models, see Amemiya 1981 and Greene 2000.

(1)

10

where Pi denotes the probability that the ith

farmer has adopted microirrigation technology(Fi = 1) and

where β0 is the intercept, βi is a slope parameterin the model, and Xi is an independent variable. Inthe logit model, like in any nonlinear regressionmodel, the parameters are not necessarily themarginal effects (Greene 2000; Kennedy 2001).They rather represent changes in the natural log ofodds ratio for a unit change in the explanatoryvariables.4 The marginal effect (or the quantitativeimportance of the explanatory variables) for thelogit model is expressed as follows:

In the present case, the variableshypothesized to influence microirrigation adoptiondecisions are summarized in table 3. Thevariables were selected based on literaturereviews of the determinants of microirrigationadoption (Caswell 1999; Shrestha andGopalakrishnan 1993; Sakks 2001), theresearchers’ own perceptions of thesocioeconomic setting of the study locations, andthe technical attributes of the microirrigationsystems prevalent in the study locations. Thesevariables may be conveniently classified into:

1. Family size and demographic structure: Thisgroup of variables includes the number ofhousehold members and the proportion ofhousehold members whose ages are lowerthan 14 years or more than 65 years (ordependency ratio). These variables indicatethe degree of labor availability in thehousehold.

2. Human capital variables such as age andlevel of education of the household head: Thevariable age may be a surrogate for manyother socioeconomic variables includingexperience or skill, wealth and conservatism.The effect of this variable on microirrigationadoption decisions depends on which ofthese age dimensions dominates. The levelof education augments extension servicesand is hypothesized to positively contribute tothe microirrigation adoption probability.

3. Ownership of agro-wells and pumps and theirtechnical attributes: The propensity of agro-well owners to adopt microirrigation systemsis expected to be higher than that of surfacewater irrigators mainly due to differences inthe type of property rights associated with thetwo modes of irrigation. Whereas surface

Description of the HypothesizedMicroirrigation Adoption Variables

As a prelude to the analyses of the logit modelresults, let us first discuss the explanatoryvariables included in the model. The mostcommon variables used in modeling technologyadoption processes are human-capital variables(e.g., level of education and age), attributes ofthe technologies, nature of the farming system asinfluenced by the interplay of various biophysicaland socioeconomic variables, tenure system,resource endowment, risk and uncertainty, socialcapital and social psychological factors (Feder etal. 1985; Rogers 1995; Legans 1979; Buttel et al.1990).

4After a few steps of transformation , equations 1 and 2 may be simplified as follows:

Here βi is interpreted as a change in the log of odds associated with a one-unit change in an explanatory variable.

(3)

(2)

11

TAB

LE 3

.D

escr

iptio

n of

var

iabl

es in

clud

ed in

the

logi

t re

gres

sion

mod

el.

Varia

ble

Mah

aras

htra

Guj

arat

Ado

pter

sN

on-a

dopt

ers

Ado

pter

sN

on-a

dopt

ers

Uni

tM

ean

SD

Mea

nS

DM

ean

SD

Mea

nS

D

Fam

ily s

ize

Num

ber

4.5

1.1

4.3

1.3

4.3

1.5

4.5

1.2

Age

of

the

hous

ehol

d he

adY

ear

44.7

12.4

45.2

14.9

43.9

12.0

45.2

11.9

Dep

ende

ncy

ratio

Per

cent

age

16.4

19.8

17.1

19.7

13.4

18.7

15.4

19.8

Year

s of

sch

oolin

g of

the

hou

seho

ld h

ead

Yea

r9.

24.

37.

54.

15.

14.

63.

94.

4

Pro

port

ion

of h

igh

cast

eP

erce

ntag

e82

.7-

52.5

-80

.0-

45.7

-

Dep

th o

f w

ell

Foo

t63

.959

.319

.236

.275

.656

.062

.642

.7

Acc

ess

to g

roun

dwat

erP

erce

ntag

e96

.3-

52.5

-10

0-

93.0

-

Pow

er o

f th

e pu

mp

owne

dH

orse

pow

er6.

612

.93.

81.

54.

012.

80.

61.

7

Sha

re o

f ce

real

s an

d pu

lses

Per

cent

age

13.8

22.0

20.5

31.4

2.8

8.4

4.1

10.6

Sha

re o

f fr

uit

crop

sP

erce

ntag

e27

.334

.55.

114

.64.

114

.37.

218

.1

Sha

re o

f ve

geta

bles

Per

cent

age

7.2

16.6

7.9

22.1

6.0

13.8

2.2

7.6

Sha

re o

f co

tton

and

oils

eeds

Per

cent

age

35.3

32.5

47.3

34.2

75.4

26.5

72.0

24.2

Sha

re o

f liv

esto

ckP

erce

ntag

e6.

412

.25.

912

.54.

18.

53.

77.

0

Sha

re o

f of

f-fa

rm a

nd n

on-f

arm

inc

ome

Per

cent

age

4.2

13.8

4.4

15.1

6.9

16.2

8.0

15.6

Pov

erty

Ind

exS

core

0.3

0.9

-0.7

0.8

0.2

0.9

-0.4

1.1

Not

e: S

D =

sta

ndar

d d

evia

tion.

12

water sources are owned publicly orcommunally, groundwater sources are usuallyowned privately. Thus the water supply fromgroundwater sources is more reliable andflexible than that from surface water sources.Hence, ceteris paribus, agro-well owners havethe motivation or incentive to use theavailable water as efficiently as possiblethrough employing various mechanisms suchas microirrigation systems. The technicalcharacteristics of agro-wells (e.g., depth) andpumps (horsepower) impinge on the owners’decision to adopt microirrigation technologiesdue to implications for energy cost or thescarcity of well water.

4. Cropping pattern or the farming system: Thisgroup of variables is represented by theshares of the different categories of cropsgrown and livestock kept in the total annualincome of the farmers. The effect of croppingpattern or the farming system on themicroirrigation adoption decisions is expectedto be substantial because of the cropspecificity of the available microirrigationsystems.

5. Other socioeconomic variables: Includedunder this group are the caste or socialstatus of the farmer, the poverty index valueof the farmer created using principalcomponent analysis, and the share of incomefrom off-farm and non-farm activities. Ceterisparibus, high caste households are expectedto have a higher probability of adoptingmicroirrigation technologies. Similarly, farmerswith a higher poverty index value (wealthierfarmers) have a higher likelihood of beingmicroirrigation technology adopters. The share

of the non-farm and off-farm income variableis also hypothesized to increase theprobability of adopting the technologiesthrough the availability of additional cash forprocuring microirrigation technologies, whichare normally considered as capital-intensive.

Transcendental Production Function

For the data generated through experimentation orcontrolled trials, marginal analysis based onpartial farm budgets is the most commonapproach used in assessing the economicadvantage of new technologies (CIMMYT 1988).This is because, in controlled trials, one or twotreatments are varied at a finite number of fixedlevels. In the case of survey data, however, alllevels of input vary continuously and hence it isdifficult to estimate incremental gains and costsat discrete points where input levels change. Inthis case, continuous analysis is used toevaluate the technical and economic efficiency ofthe different microirrigation technologies.Specifically, we fitted a transcendental productionfunction to the data to estimate marginalproductivities and measure the importantinteractions among the various yield-influencingfactors.

The transcendental production function is anextension of the well-known Cobb-Douglasproduction function and represents theneoclassical three-stage production process.5 Itcan be viewed as a generalization of the Cobb-Douglas production function that can depict thethree stages of production and has variableproduction elasticity and may be specified asfollows:

5The limitation of the Cobb-Douglas production function is that it can represent only one stage of production at a time and assumesfixed production elasticities, which requires that Average Physical Product and Marginal Physical Product are at a fixed proportion toeach other (see Debertin 1986).

(4)

13

6It was in response to this that NGOs such as IDE and Appropriate Technologies for Enterprise Creation (APPROTEC) have introducedlow-cost alternatives to the conventional capital-intensive ones.

where, X represents an agronomic andirrigation water input (continuous variables). Theseinputs are weeding cost (man-days/ha or Rs/ha),seed cost (Rs/ha), pesticide (l/ha), irrigation waterpumping (hr/ha), nitrogen (kg/ha), phosphorus (kg/ha) and potassium (kg/ha). And D is a dummyvariable representing the different microirrigationtechnologies, namely, conventional drip systems,microtube drip systems, low-cost drip systems,microsprinklers and conventional sprinklers.

The transcendental production functionspecified in equation 4 can easily be transformedto natural logs to yield

where Y is output, αi measures elasticity, Xi

measures the level of continuous variable i, γi

is the rate of change in the elasticity, γij

indicates the interaction effects between thecontinuous variables i and j (i, j = 1, 2,….., n), dj

is a measure of the effect of dummy variables ormeasures the deviation of the mean effect ofdummy variable Dl (the different kinds ofmicroirrigation technologies) from the overallmean effect (common intercept).

Optimization rules are used to defineefficiency. From the output optimization problem,optimality conditions require that the efficientregion of production is where the gain in outputper extra units of Xi is increasing:

where, MPPi is the Marginal Physical Productof input i.

The point beyond which output starts todecrease with additional units of Xi (MPPi < 0) is

where the region of inefficiency begins. Thecondition for economic efficiency is derived fromoptimality conditions for the profit-maximizingfirm, where profit (π) is computed as:

where, P is the price of output, Y(X) is theproduction function specified above, and x and Care vectors of inputs and input prices,respectively. Conditions for profit maximizationrequire that:

implying that it is economically efficient tocontinue using more units of input Xi up to thepoint where the value of the marginal product ofan extra unit of Xi (VMPi) is equal to its cost (Ci).

Principal Component Analysis

One of the hypotheses about microirrigationtechnologies is that their adoption among poorfarmers is inhibited by the high initial capitalrequirement.6 However, there is little informationregarding the adoption patterns of the differentmakes and modes of microirrigation technologiesamong the potential adopters differentiated bypoverty status. Specifically, it is necessary tosee whether the poorest section of the farmingcommunity is the prime beneficiary of the low-cost microirrigation systems as originallyenvisaged. These issues were investigated by:(1) analyzing the relative poverty status of theadopting and non-adopting farmers using what isknown as an indicator-based poverty assessmenttool (see Henry et al. 2003 and Namara et al.2003), (2) grouping the farmers into five

(5)

(6)

(7)

(8)

14

Dynamics of Microirrigation Adoption

There is a marked difference in the temporaladoption pattern of microirrigation technologiesbetween Maharashtra and Gujarat. TheMaharashtra farmers in the sample were awareof the technologies since the early 1980s,while in Gujarat farmers knew about thetechnologies only since the early 1990s (figure3). However, the peak period of adoptionoccurred within the last 5 or 6 years in both

states (figure 4). In both locations, about 90percent of the non-adopters are aware ofmicroirrigation technologies.

In Gujarat, there is a contrast regarding theinformation source of microirrigation betweenadopters and non-adopters (table 4). The currentmicroirrigation adopters initially learned about itfrom diverse sources including NGOs, otherfarmers and public extension systems, in thatorder of importance. The non-adopters had arelatively limited range of sources of information

poverty strata (i.e., very poor, poor, middle,rich and very rich), and (3) analyzing thepattern of distribution of the adopters ofdifferent kinds of microirrigation technologiesacross the five identified poverty strata.

The indicator-based poverty assessmenttool first identifies the strongest individualindicators that distinguish relative levels of

poverty, such as human resources (e.g., yearsof schooling, age, etc.), dwellingcharacteristics and assets (e.g., farm andhousehold assets, etc.). Second, the methodpools the explanatory power of the identifiedindividual indicators into a single index usingprincipal component analysis.

FIGURE 3.Year of first-time awareness among farmers about microirrigation technologies.

Analyses of Microirrigation Adoption

15

TABLE 4.Sources of information for microirrigation dissemination.

No. Source Gujarat Maharashtra

Adopters Non-adopters Adopters Non-adopters

(%) (%) (%) (%)

1 NGOs 35.3 33.3 0.0 0.0

2 Other farmers 26.1 43.9 84.7 85.8

3 Dealers 7.8 4.5 0.0 0.0

4 Cooperatives 5.2 6.1 0.0 0.0

5 Government/public extension 11.1 1.5 0.0 2.9

6 Company/manufacturer 5.2 0.0 9.8 0.0

7 Private business people 3.9 0.0 0.0 0.0

8 Village agrifair 1.3 0.0 0.0 0.0

9 Banks 0.7 0.0 0.0 0.0

10 Agents 1.3 0.0 1.3 0.0

11 Newspapers and TV 0.0 1.5 0.0 0.0

N 152 60 156 35

FIGURE 4.Year of first practical use of microirrigation technologies.

and a significant proportion of them heard aboutmicroirrigation technologies from other farmerseither by word of mouth or through fieldobservation. In Maharashtra, there is nosignificant difference between adopters and non-

adopters regarding the information source formicroirrigation. Both categories became aware ofthe technologies from other farmers. This may bedue to the fact that in this area the technologyhad already taken root.

16

The farmers in Gujarat noted about 19advantages of microirrigation technologies whilethose in Mharashtra mentioned 15 advantages(table 5). These show that the adoption ofmicroirrigation technologies has complex physical,biological and social ramifications. In bothlocations water and power savings wereconsistently mentioned by the majority of thefarmers as the most important merits ofmicroirrigation technologies.

In Maharashtra, about 97.5 percent of thefarmers who are currently using microirrigationtechnologies have indicated that they willcontinue using the technologies. Thecorresponding figure for Gujarat is 96.7 percent. Afew farmers from both states intend todiscontinue the use of microirrigationtechnologies. The main reasons given fordiscontinuance are summarized in table 6. Some

farmers also claimed that there was nosubstantial benefit from microirrigation technologyadoption or that the flooding method did givebetter results. Frequent rodent damage and lackof suitable land (i.e., infertile and stony fields) aresome of the other adoption constraining factors.

Similarly, the farmers who intend to continueusing microirrigation technologies were askedwhether they planned to increase the area undermicroirrigation. About 80 percent of the farmers inGujarat and 30.8 percent in Maharashtraresponded that they would like to put more landunder microirrigation. The farmers who were notwilling to increase the area under microirrigationgave the following reasons: (1) shortage of land,(2) water shortage, (3) microirrigation systems arenot durable, (4) limited capacity of the pumpsthey own, (5) salinity problems, specifically inGujarat, (6) insufficient subsidy (reduced from an

TABLE 5.Perceived advantages of microirrigation technologies.

No. Advantage Gujarat farmers (%) Maharashtra farmers (%)

1 Saves water 96.1 98.8

2 Saves power 65.4 96.9

3 Reduces disease incidence 60.0 0.0

4 Higher yield per unit area 35.9 0.6

5 Convenient irrigation timing 28.1 77.3

6 Reduces labor costs 25.5 21.5

7 Allows irrigated area expansion 24.8 4.9

8 Improves quality of produce 24.2 6.7

9 Extends irrigation time 23.5 41.7

10 Even distribution of water in the field 23.5 25.4

11 Allows pre-monsoon planting 20.9 3.1

12 Reduces weed growth 20.9 0.0

13 Reduces crop failure 18.3 0.0

14 Can work on undulating fields 18.3 4.9

15 Reduces soil erosion 13.1 0.0

16 Enables crop production in summer 12.4 6.1

17 Allows efficiency in manure use 11.1 1.8

18 Allows efficiency in fertilizer use 9.8 2.5

19 Reduces insect damage 1.3 2.5

N 153 156

17

initial 66 percent to 25 percent), (7) lack ofsuitable microirrigation technology, and (8)maintenance difficulties.

Factors Influencing the Adoption ofMicroirrigation Technologies

Having probed into the mechanics of thehypothesized relationships between the dependentand explanatory variables, we now turn to theanalyses of the results of the logit adoptionmodel (table 7). Most of the variables included inthe model had the expected signs. Family sizeand dependency ratio (i.e., the proportion offamily members whose ages are less than 14 ormore than 65), which were included to indicatethe status of labor availability in the household,had an insignificant effect reflecting the lowerlabor requirement of microirrigation technologiesas compared to the traditional irrigation methods.

The human-capital variables are of specialinterest in relation to the microirrigation adoptionprocess because of the fact that thesetechnologies need technical and managerial skillfor proper utilization. The age variable had a

positive but insignificant effect on adoptionprobability in both Gujarat and Maharashtra.Normally one would expect the older farmers tohave a lower chance of adoption of newinnovations (Neil and Lee 2001). The observedunexpected sign of age in the present analysismight be because the age variable may havecaptured more the experience and wealth aspectsthan its conservatism dimension. The interactionbetween education and age had the expectedsign, but was statistically significant only inGujarat. The reason for the insignificance ofthese variables in Maharashtra may be thatmicroirrigation systems have a long history andare no more considered as new technologies inthis area. In other words, microirrigationtechnologies are used by most farmers. Thenegative sign for the years of schooling-by-ageinteraction effect could be understood from thecommonly observed negative correlation betweenthe age and level of education. In other words,younger farmers tend to be more educated thantheir older counterparts.

As we anticipated, the ownership of dug wellsand bore wells had a strong effect on theprobability of adoption of microirrigationtechnologies in both states. This is because well

TABLE 6.Perceived disadvantages of microirrigation adoption.

No. Disadvantage Gujarat farmers (%) Maharashtra farmers (%)

1 High initial capital requirement 40.5 1.8

2 Clogging problems 36.6 0.0

3 Maintenance problems 30.7 3.1

4 Difficult to install 25.5 3.1

5 Requires know-how and skills 18.3 0.6

6 Crop specificity 16.3 0.0

7 Limits crop diversification 14.4 0.0

8 Interferes with harvesting 11.8 0.0

9 Increases labor demand 7.2 0.0

10 Interference with agronomic practices 5.9 0.0

11 Tail-end plants get less water 4.6 0.0

12 Cumbersome procurement process 3.9 0.0

N 153 163

18

TAB

LE 7

.F

acto

rs i

nflu

enci

ng t

he a

dopt

ion

of m

icro

irrig

atio

n te

chno

logi

es.

Varia

ble

Guj

arat

Mah

aras

htra

Coe

ffici

ent

t-ra

tioM

argi

nal

effe

ctC

oeffi

cien

tt-

ratio

Mar

gina

l ef

fect

Con

stan

t-6

.370

-1.7

48a

-0.9

9841

-7.0

481

-2.4

08b

-0.4

367

Fam

ily s

ize

0.20

791.

176

0.03

258

0.27

011.

266

0.01

673

Age

of

hous

ehol

d he

ad-0

.000

09-0

.004

-0.0

0001

0.04

381.

164

0.00

271

Year

s of

sch

oolin

g of

hou

seho

ld h

ead

0.26

821.

294

0.04

203

0.24

701.

071

0.01

531

Year

s of

sch

oolin

g-by

-age

int

erac

tion

-0.0

045

-0.9

64-0

.000

70-0

.006

3-1

.250

-0.0

0039

Cas

te1.

0447

2.43

0b0.

1637

31.

1903

2.42

0b0.

0737

5

Dep

ende

ncy

ratio

-0.0

0045

-0.0

35-0

.000

07-0

.010

9-0

.846

-0.0

0068

Pov

erty

ind

ex0.

5569

2.36

6b0.

0872

80.

5385

1.60

0a0.

0333

7

Acc

ess

to g

roun

dwat

er3.

1345

1.10

10.

4912

64.

4406

3.70

7c0.

2751

4

Dep

th o

f w

ell

0.00

200.

459

0.00

031

0.00

681.

034

0.00

042

Hor

sepo

wer

of

pum

p0.

5342

5.59

0c0.

0837

20.

3258

2.72

2c0.

0201

9

Sha

re o

f ce

real

s an

d pu

lses

-0.0

545

-1.7

61a

-0.0

0855

-0.0

156

-0.9

43-0

.000

97

Sha

re o

f fr

uit

crop

s0.

0032

00.

144

0.00

0501

0.05

075

2.23

2b0.

0314

5

Sha

re o

f ve

geta

bles

0.05

652.

027b

0.00

8860

-0.0

195

-1.0

35-0

.001

21

Sha

re o

f co

tton

and

oil

crop

s0.

0110

90.

613

0.00

1738

-0.0

067

-0.4

17-0

.000

41

Sha

re o

f liv

esto

ck0.

0199

10.

653

0.00

3120

0.05

925

1.90

5a0.

0036

7

Sha

re o

f of

f-fa

rm a

nd n

on-f

arm

inc

ome

-0.0

2189

-0.4

15-0

.003

430.

0890

52.

598c

0.05

518

Squ

are

of s

hare

of

off-

farm

and

no

n-fa

rm i

ncom

e0.

0006

00.

537

0.00

0094

-0.0

0192

-3.0

24c

-0.0

0012

Logl

ikel

ihoo

d fu

nctio

n-8

2.79

477

-63.

2210

7

2χ

(d.

f.)11

4.21

33(1

7)c

135.

2486

(17)

c

Per

cent

age

of c

orre

ct p

redi

ctio

ns83

.986

.1

a Sig

nific

ant

at 1

0%

b S

igni

fican

t at

5%

c Sig

nific

ant

at 1

%

19

owners have a high degree of control over thewater source and have the motivation toefficiently use the water available in the well.Moreover, the depth of well (in Gujarat) and thehorsepower of the pump owned (in Maharashtra)had a significant positive impact on the likelihoodof adopting these technologies. This indicatesthat as the wells get deeper, the farmers areobliged to spend more money for pumping.Therefore, the higher adoption probability amongfarmers owning deeper wells and higherhorsepower pumps partly reflect the farmers’power-saving motives.

Contrasting results were found forMaharashtra and Gujarat regarding the effect ofthe cropping pattern on the adoption ofmicroirrigation technologies. In Maharashtra, theshare of fruit crops in the cropping system had asignificant effect on adoption probability, while inGujarat the share of vegetables, cotton and oilseeds (mainly groundnut) had a positive influenceon adoption probability. The model also showsthat in both states the higher the share of cerealsand pulses in the cropping pattern the lower theprobability of adoption of microirrigationtechnologies. However, this relationship was notstatistically significant in Maharashtra. Thesefindings are quite in line with the observation ofactual conditions in the study areas. InMaharashtra, farmers mainly use microirrigationtechnologies for cultivating fruit crops such asbanana and grape, while in Gujarat they use itlargely for groundnut, cotton and vegetables.

The effect of cropping pattern-cum-farmingsystem on the microirrigation adoption processshould be evaluated on short- and long-termperspectives. In the short run, as this studyindicates, microirrigation systems may bypasscertain groups of farmers due to their inherentcrop specificity. The bulk of the availablemicroirrigation technologies is suitable for fruitcrops, vegetables and cash crops such as cottonand groundnut. Therefore, farmers growing staplefood crops such as cereals and pulses are notsufficiently benefiting from innovations inmicroirrigation technology. In the long run,

however, this scenario may be reversed due totwo possible developments: (1) microirrigationengineers may innovate systems to irrigate cropsfor which the currently used microirrigationapplications are not suitable (Polak et al. 1997),and (2) farmers may shift their cropping patternsto benefit from innovations in microirrigationtechnology. The latter response will havesubstantial impacts on the water resourceseconomy of a watershed or basin and also oncash crop prices.

All the socioeconomic variables, namely,membership in a high caste group, poverty indexand share of income from off-farm and non-farmactivities, in order of quantitative significance(see the respective marginal effect values in table7), had significant impacts on the decision toadopt microirrigation technology. On average,well-to-do farmers are most likely to adoptmicroirrigation technologies in both states. An in-depth analysis of the poverty outreach of differentmicroirrigation technologies is presented in thesection on Degree of Poverty Outreach ofAlternative Microirrigation Technologies (page 25).

In Maharashtra, the share of off-farm andnon-farm income and its square had the expectedsign. As the farmers’ share of income from off-farm and non-farm sources increases, thelikelihood of adopting microirrigation technologiesincreases—but only up to a certain point. Thisshows the importance of cash in the initialadoption decision of farmers. However, at higherlevels of the share of off-farm and non-farmincome, the farmers are less likely to adoptmicroirrigation technologies because agricultureloses its importance and is no longer theprimary source of livelihood. The situation inGujarat is the exact opposite of that inMaharashtra. In the Gujarat case, as the shareof income from off-farm and non-farm sourcesincreases the likelihood of adoptingmicroirrigation technologies decreases up to acertain point and then increases. This may bethe reflection of the differences in the nature ofoff-farm and non-farm activities between thetwo study locations.

20

This section presents the results of agro-economic analyses (i.e., yield responses, andtechnical and economic efficiency of agronomicand irrigation technology inputs) and povertyimpact analyses (i.e., the poverty outreach ofmicroirrigation technologies) and discusses theresultant implications for extension, povertyreduction and sustainable water (groundwater)management. The estimated results of thetranscendental production function fitted to datato assess the yield responses of the differentagronomic and water management inputs arepresented in table 8. The results of the technicaland economic efficiency analyses are presentedin tables 9 through 11 and figures 5 to 8 depictthe extent of poverty outreach of microirrigationtechnologies.

Yield Responses

In both study locations, crop production basedsolely on microirrigation use is rarely found. Foradopters, microirrigation use is oftencomplemented with a flooding/furrow method ofirrigation at least once during the croppingseason. Farmers use microirrigation technologies:(1) to enable early planting (e.g., cotton andgroundnut) so that plants are already establishedat the time of the onset of rain during themonsoon and make efficient use of rainwater, (2)to safeguard crops against crop loss or yieldreduction due to a dry spell or early withdrawal ofrain, and (3) to save groundwater for use duringthe summer and Rabi seasons.

The coefficient for a variable in the naturallog form, αi, measures the elasticity of the outputto the use of the input, while the coefficient for avariable in levels (γi ) measures the rate ofchange in the elasticity of production.

Among the agronomic inputs, the highestbanana yield response was observed for pesticideapplication followed by that for potassium. Thecoefficient for the banana yield response tophosphorus suggests that the farmers are

currently applying this input above the normallyrequired rate. For groundnut, the highestresponses were observed for weeding and seedcost, respectively. However, the groundnut yieldresponses to agronomic inputs are notstatistically significant. The coefficient for thelevel of irrigation water use (as indicated bypumping hours per hectare) in groundnutproduction is positive both in the actual level andlog form. This indicates that the farmers arecurrently applying an insufficient level of irrigationwater to groundnut. Most of the agronomic inputsincluded in the cotton yield response function arestatistically significant. Among these, the highestyield response was observed for seed inputfollowed by that for nitrogen.

The transcendental yield response functionindicates that for all of the three cropsconsidered (banana, groundnut and cotton), theuse of microirrigation technologies generallyresulted in a significant yield improvement overthe traditional irrigation practices (see theestimates of the dl coefficient in table 8). Forbanana and cotton, yield response to low-costdrip irrigation is lower than that for conventionaldrip. For groundnut, the response to sprinklerirrigation is by far better than the response to dripirrigation. Dhawan (2002) also reported similarobservations. In addition, farmers in the studyarea (in Gujarat) claim that the use of driptechnology for groundnut is marred by manytechnical problems.

The synergistic effect of the variousagronomic inputs can be observed from thepositive interaction effect coefficient (γij ).Pesticide-by-seed interaction effect is significantfor banana and cotton. This shows that diseasesand pests are important production constraints forthese crops and that the joint yield effect ofthese inputs is more than the sum of theirindependent effects. Similarly, the significantseed-by-weed interaction effect for cotton impliesthat the response to the weeding efforts of thefarmer depends on the weed competitiveness andyield potential of the cotton variety used.

Agro-economic Analyses of Microirrigation Technologies

21

TABLE 8.

Results of the transcendental regression model.

No. Variable Banana Groundnut Cotton

Coefficient t-ratio Coefficient t-ratio Coefficient t-ratio

1 Constant 3.172 1.408 -1.097 -0.630 -5.158 -3.635c

2 Estimates of iγ (agronomic and water management input in levels)

Weeding -0.004362 -0.831 -0.0001653 -1.251 -0.01429 -1.169

Seed -0.0000331 -0.633 0.00000424 0.088 -0.0005785 -3.820c

Pesticide -0.378 -2.121b -0.01156 -0.282 -0.05139 -1.500

Pumping NA NA 0.001147 0.516 -0.0007448 -0.612

Nitrogen -0.001618 -0.985 -0.002253 -0.908 -0.006425 -3.054

Phosphorus 0.003645 1.685a NA NA -0.002535 -0.474

Potassium -0.003698 -1.251 NA NA NA NA

3 Estimates of iα (agronomic and water management input in natural log form)

Weeding 0.485 2.324b 0.175 1.319 0.0782 0.348

Seed -0.08461 -0.38 0.237 1.177 0.612 3.501c

Pesticide 0.647 1.343 0.08841 0.713 0.321 2.743c

Pumping NA NA 0.05303 0.410 0.225 2.432b

Nitrogen 0.196 0.63 0.03285 0.218 0.593 2.670c

Phosphorous -0.524 -1.589 NA NA 0.04591 0.194

Potassium 0.575 1.854a NA NA NA NA

4 Estimates of ld (microirrigation technology)

Conventional drip 1.214 3.138c NA NA 1.162 3.516c

Low-cost drip 1.05 2.997c NA NA 0.911 3.214c

Microtube drip NA NA 0.511 1.787a 0.627 3.779c

Microsprinkler NA NA 0.650 2.876c NA NA

Conventional sprinkler NA NA 0.591 2.012b NA NA

5 Interaction ijγ (among selected agronomic and irrigation water input variables)

Pest by seed 0.0000124 1.954a NA NA 0.0000116 1.690a

Seed by weed 0.0000003 0.478 0.00000001082 1.308 0.0000071 2.444b

Pest by weed NA NA NA NA 0.0003578 0.878

Pump by seed NA NA NA NA 0.0000005 0.594

Nitrogen by potassium 0.0000066 1.041 NA NA NA NA

Adjusted R2 0.277 0.108 0.391

F Statistic 2.687c 2.545c 4.848c

Durbin-Watson Statistic 1.753 1.832 1.882

N 76 180 115

a Significant at 10% b Significant at 5% c Significant at 1%

Note: NA = Not available or not applicable.

22

Technical and Economic Efficiency

The technical and economic efficiencyparameters, i.e., Marginal Physical Product(MPP) and Value of Marginal Product (VMP), fordifferent microirrigation technologies, derived fromthe fitted transcendental response functionsdiscussed in the preceding section, are presentedin table 9.7 The MPP values indicate an extrayield advantage that a farmer obtains whenshifting from the traditional irrigation methods to amicroirrigation practice. MPP values shown intable 9 indicate that the use of microirrigationtechnologies in banana, groundnut and cottoncultivation is technically efficient (MPP > 0) andthat except for groundnut the technical efficiencyof conventional drip systems is superior to that oflow-cost drip systems. The superior yieldperformance of conventional microirrigationsystems relative to the low-cost systems may atfirst glance contradict the observed occurrence ofboth systems at the same time in a given area oreven in a given farm. However, these two

systems may serve different purposes and aretherefore not mutually exclusive. Low costmicroirrigation technologies are primarily adoptedto save crops, including perennial crops such asorchards and coconut, during drought years byextending the use of the limited water as muchas possible.

The yield advantage of the use of dripsystems in groundnut cultivation is lower thanthat of the sprinkler systems. This result confirmsthe claim by farmers that drip is normally notpreferred for groundnut irrigation. The choicebetween drip and sprinkler irrigation systems isalso influenced, in addition to expected yield, bythe quality of groundwater. The farmers indicatedthat the use of sprinkler irrigation system in areaswhere the water is saline is quite risky as it mayaffect plant growth. It was in recognition of thisrisk that AKRSP instituted two different subsidyregimes for farmers using sweet water and thoseusing saline water. These two categories offarmers are entitled to a 33 percent and a 50percent subsidy, respectively.

TABLE 9.Technical and economic efficiency of microirrigation systems.

Crop Microirrigation MPP VMP Investment Subsidized

technology (q) (Rs) cost (Rs) investment cost (Rs)

Banana Low-cost drip 142.2 42,659 27,360 13,680

Conventional drip 181.2 54,353 55,000 27,500

Groundnut Microsprinklers 6.96 10,301 22,239 11,120

Microtube drip 4.35 6,440 29,652 14,826

Conventional sprinklers 5.21 7,706 30,000 15,000

Cotton Low-cost drip 7.26 11,916 10,081 5,041

Microtube drip 4.99 8,201 17,087 8,544

Conventional drip 9.26 15,199 45,825 22,913

Notes: Rs = Indian rupee; US$1.00 = Rs 43.70.

q = quintal; 1 quintal = 100 kilograms.

The price of banana was estimated at Rs 300 per quintal.

MPP = Marginal Physical Product; VMP = Value of Marginal Product.

7See Appendix 2 for details of how these MPP values were derived from the fitted transcendental response functions.

23

The main motive of farmers in usingmicroirrigation technologies, particularly forgroundnut, is to plant crops earlier, about a monthearlier than the normal practice. They do this forone or more of the following reasons:

• To take advantage of the available water sothat they can spread water over a largerextent and establish the crop by the time themonsoon sets in.

• To be able to harvest a normal crop even incase of early withdrawal of the monsoon.

• To create a better soil moisture conditionthroughout the crop season to obtain higheryields and better output quality as a result ofbetter pod filling.

• To enable early harvesting, thus reducinglabor costs.

However, in case the rain continues duringSeptember (harvest time), the quality of theproduce may be adversely affected. The otherrationale for adopting microirrigation in groundnutis to save water for use in the cultivation of pearlmillet and vegetables in summer.

Technical efficiency alone does not guaranteeeconomic efficiency. A farmer may well operate inthe technically efficient region of the productionfunction but may still be judged as economicallyinefficient, based on considerations of input-output price relations. Thus, to evaluate theeconomic efficiency of the different microirrigationtechnologies we need to consider the input-outputprice relationships. In the present case, this wasachieved through calculating the VMP for each ofthe microirrigation technologies and comparing itwith their respective initial investment costs undertwo scenarios, i.e., actual costs and subsidizedcosts (see table 9).

From the results of the economic efficiencyanalyses the following inferences may be made:

• Even under the very conservative scenario ofcomparing the VMP with the actualinvestment cost,8 except for microtube dripand conventional sprinklers under groundnut,all of the microirrigation technologies areeconomically efficient and the farmers canrecuperate their initial investment capitalwithin 1 to 3 years.

• Subsidies further increased the profitability ofinvestments in microirrigation technologies.

• The magnitude of economic gains frominvestments in microirrigation technologiesdepends on the type of crop. Microirrigationtechnology use in banana is highlyremunerative, and cotton comes second inthis respect. The VMP for banana is almostequal to the initial investment cost. Thismeans that the farmers can recuperate theinvestment cost within 1 or 2 years of use.

• An investment in conventional drip systemsis economically more rewarding than that inlow-cost drip systems.

One of the advantages of microirrigationtechnologies recognized by the sample farmers inthis study, and often also claimed by those inmany other similar studies, is that they enhancethe productivity of other agronomic inputs inaddition to that of water. To investigate theseeffects we fitted a separate transcendentalresponse function to microirrigation technologiesand traditional irrigation methods for groundnutand cotton and calculated the MPP and VMPvalues (see tables 10 and 11). The results forgroundnut are entirely consistent with ourexpectations in that the calculated MPP and

8 Ideally the VMP figures ought to be compared to the annual ownership cost of microirrigation technologies, which obviously is afraction of the initial investment cost.

24

VMP values under sprinkler irrigation systems arehigher than that for traditional irrigation methods.Thus, microsprinklers do enhance the marginalproductivity of other agronomic inputs. However,the farmers are applying more water and nitrogento groundnut under the traditional method ofirrigation than economically justifiable.

The results for cotton are mixed in thatcontrary to the expectations the MPP figures forseed, phosphorus and pesticide are higher for

traditional irrigation methods than that under dripirrigation systems. However, there are justifiablereasons for these anomalies. For instance,farmers claim that with the drip method ofirrigation cotton plants are not able to make fulluse of DAP applied as basal fertilizer and that inmost cases the fertigation tank attached to thesystem is used for applying only liquid urea. Theexplanation for the unexpected results of seedand pesticide inputs is related to the higher

TABLE 10.Technical and economic efficiency of other agronomic inputs for groundnut under different irrigation methods.

Agronomic input Unit input Traditional methods Sprinkler systems

price (Rs) MPP (q) VMP (Rs) MPP (q) VMP (Rs)

Nitrogen (kg) 10.50 -0.014238074 -21.1 0.02554 37.8

Pesticide (l) 252.0 0.233088759 345.0 0.60806 899.9

Weeding cost (Rs) 1 0.001738573 2.6 0.00233 3.4

Seed cost (Rs) 1 0.000638345 0.9 0.00107 1.6

Pumping (hours) 32.5 0.014114507 20.9 0.03340 49.4

Notes: Rs = Indian rupee; US$1.00 = Rs 43.70.

q = quintal; 1 quintal = 100 kilograms.

The price of groundnut was estimated at Rs 1,480 per quintal.

MPP = Marginal Physical Product; VMP = Value of Marginal Product.

TABLE 11.Technical and economic efficiency of other agronomic inputs for cotton under different irrigation methods.

Agronomic input Unit input Traditional methods Drip systems

price (Rs) MPP (q) VMP (Rs) MPP (q) VMP (Rs)

Nitrogen (kg) 10.5 -0.0103657 -17.0 0.001102 1.8

Phosphorus (kg) 16.2 0.0601041 98.7 0.016049 26.4

Pumping (hours) 32.5 0.0135367 22.2 0.036285 59.6

Weeding (man-days) 50.0 -0.1495965 -245.6 -0.10011 -164.4

Pesticide (l) 252.0 0.7827098 1285.2 0.541208 888.7

Seed cost (Rs) - 0.0031636 5.2 0.001096 1.8

Notes: Rs = Indian rupee; US$1.00 = Rs 43.70

q = quintal; 1 quintal = 100 kilograms.

The price of cotton was estimated at Rs 1,642 per quintal.

MPP = Marginal Physical Product; VMP = Value of Marginal Product.

25

adoption rate of the Bt cotton9 variety amongfields irrigated through traditional methods(flooding). The Bt cotton variety is resistant to thebollworm (enabling savings in pesticide costs)and is high yielding. Similar to the case forgroundnut, for cotton too the farmers tend toapply more water and nitrogen than economicallyjustifiable under traditional irrigation methods.

In summary, it is shown that the use ofmicroirrigation in the cultivation of banana, cottonand groundnut is both technically andeconomically justifiable (i.e., higher marginalphysical productivities with the value of marginalphysical products far greater than the resourcecost). The investment costs can be recoveredwithin just one or two years of use, particularlygiven the generous subsidy schemes. Thismotivates farmers to invest more in groundwaterdevelopment. This argument is substantiated bythe fact that farmers in the two study locationsoften sink wells close to each other creatingfierce competition with subsequent stress on theaquifer resources. However, economics, thoughimportant, is not sufficient to explain the farmers’adoption behavior.

Degree of Poverty Outreach ofAlternative MicroirrigationTechnologies

The creation of a poverty index assigns to eachhousehold a poverty ranking score. The lower thescore, the poorer the household relative to allothers with higher scores. The scores formicroirrigation adopters and non-adopters can becompared to indicate the extent to whichmicroirrigation technologies reach the poor. To dothis, an appropriate definition of the poor has tobe adopted. Since the target of NGOs engaged inthe development and dissemination ofmicroirrigation technologies (such as IDE andAPPROTEC) is the poorest section of the farming

population in India, the sample farmers weredivided into quintiles (i.e., very poor, poor, middle,rich and very rich). First, the non-adopters wereranked based on their relative poverty score tocreate five poverty groupings. This is becausethe non-adopters represent the unbiased sampleof the general population. Then the cutoff valuesfor the quintiles of non-adopters were used togroup the sample of microirrigation adopters.

The poverty outreach of microirrigationtechnologies was assessed by drawing bargraphs of the distributions of the microirrigationadopting and non-adopting samples by povertyquintiles and then visually inspecting the patternof distribution (see figures 5 and 6). By default,the bars for non-adopters are expected to beequal in size across the poverty groups and, ifmicroirrigation technologies are poverty-neutral,the distribution for adopters is expected to followa similar pattern to that of non-adopters. However,this is not true in the present case.

In Gujarat, the current microirrigation adoptersare somewhat evenly distributed among themiddle, rich and very rich groups (figure 5). Asfigure 6 shows, currently the largest proportion ofmicroirrigation technology adopters in Maharashtrabelongs to the relatively very rich group. Theslight difference in the pattern of poverty-microirrigation adoption interaction betweenGujarat and Maharashtra may be because manyNGOs are operating in Gujarat. However, in bothMaharashtra and Gujarat, the poor and the verypoor categories are the least represented.

Thus, the direct poverty impact of thetechnologies is not substantial at the moment.However, the different microirrigationtechnologies have different levels of directpoverty impacts (see figures 7 and 8). Thiscan be visualized by determining how much thebar graphs for the different kinds ofmicroirrigation technologies deviate from that ofthe traditional irrigation methods. The mostcommonly reported view is that low-cost

9Bt cotton is a genetically modified seed, created by inserting a gene from Bacillus thuringiensis (Bt), a naturally occurring soil bacterium,so that the plant produces Bt toxins that kill bollworms.

26

FIGURE 6.Poverty outreach of microirrigation technologies in Maharashtra.

FIGURE 5.Poverty outreach of microirrigation technologies in Gujarat.

microirrigation technologies (low-cost drip suchas pepsee, easy drip, and microsprinklers) areeasily accessible to poor farmers by virtue oftheir low capital requirements. However, thisstudy shows that the very rich farmers

represent the highest proportion of low-costdrip technology adopters in both Maharashtraand Gujarat. In the case of Gujarat, none ofthe very poor farmers have yet accessed low-cost drip technologies.

27

FIGURE 7.Poverty outreach of different microirrigation technologies in Maharashtra.

FIGURE 8.Poverty outreach of different microirrigation technologies in Gujarat.

28

Changes in Cropping Pattern

The shifts from the traditional flooding/furrowmethod of irrigation to the use of microirrigationhave multifaceted impacts both at the householdlevel and at the higher scale of spatial andsocioeconomic aggregation. One remarkableeffect in the study locations has been thesignificant change in the cropping pattern thatoccurred following the adoption of microirrigationtechnologies (table 12). It can be seen that thetwo locations greatly differ in cropping patternirrespective of the microirrigation adoption statusowing to differences in the agro-ecologicalsettings.

A close scrutiny of the data displayed in table12 reveals the following main points regarding thecropping pattern difference between adopters andnon-adopters in Gujarat.

• The proportion of microirrigation adoptersgrowing cotton and vegetables is significantlyhigher than that of non-adopters.

• Lower proportions of microirrigationadopters cultivate cereals than non-adopters. In addition, the types of cerealgrown by the two groups of farmerssignificantly differ. Adopters grow mainlywheat, while the non-adopters grow thetraditionally drought-tolerant cereals suchas pearl millet and sorghum.

• A slightly higher proportion of farmers in thenon-adopter category grows fruit crops thanthe adopters. However, the types of fruit cropcultivated by the two groups are different.Microirrigation adopters produce high-valuebut water-intensive fruit crops such asbanana, while the major fruit crop grown bynon-adopters is coconut.

• Microirrigation adopters cultivate more diversecrops during the year than the non-adopters.

Almost similar situations were observed inMaharashtra. The following points may be maderegarding the cropping pattern differential betweenadopters and non-adopters in Maharashtra.

• Unlike the case for Gujarat, there is nosignificant difference in the proportion ofadopters and non-adopters growingvegetables.

• The proportion of microirrigation adopters inMaharashtra that grow high-value fruit crops,such as banana, is significantly higher thanthat of non-adopters. The average bananaarea cultivated is also significantly higher formicroirrigation adopters.

• Similar to the case of Gujarat, microirrigationadopters in Maharashtra cultivate morediverse crops within a year than non-adopters.

Other Impacts of Microirrigation Adoption

TABLE 12.Comparison of the cropping patterns of microirrigation adopters and non-adopters.

No. Crop Gujarat Maharashtra

Adopters (%) Non-adopters (%) Adopters (%) Non-adopters (%)

1 Groundnut and other oil seeds 54.7 63.7 1.2 7.1

2 Cotton 20.1 6.7 31.1 48.8

3 Cereals 9.7 15.5 28.7 25.0

4 Fruit crops 7.6 10.3 25.0 3.6

5 Vegetables 6.0 2.9 4.8 4.8

6 Sugarcane 0.9 0.7 0.8 1.2

7 Pulses 0.3 0.0 8.2 9.6

29

The adoption of microirrigation technologieshelped some farmers to grow Rabi and summercrops. Figure 9 depicts the trend of the areaunder microirrigation by farmers in the sampledifferentiated by season and location. It can beobserved that the microirrigation area has sharplyincreased during the 5 years from 1999 to 2003 inboth Maharashtra and Gujarat, particularly duringthe Kharif season. Similarly, an increasing trendin the microirrigation area can be observed infigure 9 during the Rabi and summer seasons forMaharashtra farmers. In Gujarat, the change inthe microirrigation area during Rabi and summerseasons is not so pronounced.

In summary, microirrigation adoption in thetwo study locations (1) led to the prominence ofhigh-value or water-intensive crops in the farmingsystem, and (2) further improved the croppingintensity. There is a tendency towards thegreening of farmlands year round resulting inmore evapotranspiration. This obviously will haveimplications for the sustainable use ofgroundwater resources.

Impacts on Women

The impact of microirrigation adoption on ruralwomen varies, depending on which of thefollowing socioeconomic categories they belongto: (1) women of landless households, (2) womenof small cultivator households, and (3) women oflarge cultivator households. Moreover, it alsodepends on the kind of the system adopted.Hence, any generalization of the impact ofadoption on women across the board would beinaccurate. Our study suggests a mixed reactionfrom various adopters.

Our data suggest that women of both studyregions adopted drip kits (like drum and bucket)either to generate additional income or to improvenutritional intake. These adopters belonged tosmall cultivator households. Women were foundto be involved in the operation and use of the kitsfor vegetable farming in homesteads. Therefore,the adoption does not necessarily reduce theirworkloads; the workloads increase because theyhave to carry out all the farming activities,

FIGURE 9.Area under microirrigation during 1999-2003.

Note: 1.00 acre = 0.40468 hectare.

30

sometimes including even marketing. However,the additional employment opportunity offered bythe technology at the homesteads saves traveltime and the privation that often accompaniesworking away from home. This is significant forwomen as they can work without fear ofharassment or humiliation and can take care ofyoung children at the same time. Moreover,household nutritional intake has significantlyimproved as the families have started eatinggreen vegetables regularly. Our qualitativediscussion revealed that women’s relativeaccess to income had also improved as theyreceived the revenue from the sale ofvegetables.

In the study locations, women provide asubstantial share of the labor input forintercropping, harvesting and irrigating duringdaytime. Besides, women perform all thehousehold chores and agricultural activities,except physically demanding work like the use ofheavy equipment. Men are more involved inprocurement, marketing and the financialmanagement of agricultural activities.

Our qualitative data suggest that women oflandless households were not able to benefitmuch, in both study regions. These womenusually work as wage earners. The reduction inthe availability of weeding jobs due to theintroduction of microirrigation, particularlycustomized systems, reinforced the vulnerabilityof women of the landless category as thesewomen found a significant decline in their laborrequirement due to lesser growth of weeds.Microirrigation technology, on the one hand,reduces the number of hours or days required forweeding while, on the other, it diminishes theearning opportunities. The situation deterioratesfurther when these women have to take theoverall burden of raising children and sustainingthe household due to the out-migration of malemembers, an increasingly common phenomenonin rural areas.

Women of households owning a customizedsystem, who are large cultivators in the surveyedlocations, benefited from the technology as it ledto a reduction in the labor requirement. Thedecline in labor inputs also helped save cash that

such families used to pay to female labor forweeding. Bilgi (1999) reported similar findingsfrom Aurangabad and Bijapur. Those householdsthat used hired labor for irrigation and otheragronomic activities realized significant moneysavings. A lesser labor requirement implies theavailability of more leisure time to share withthe family or to utilize in other productiveactivities.

In summation, the category of women thatbenefited most from the adoption ofmicroirrigation technologies was small cultivators,as they could improve their own access toincome and their family nutritional intake. Thelandless category was negatively affected due tothe reduction of the labor requirement caused bythe technology. And, women of large cultivatorhouseholds found it beneficial as their workloadwas reduced.

With regard to the impact of adoption onwomen’s household decision making, datasuggest that there was no significant changewhere women of landless and large cultivatorhouseholds were concerned. However, thedecision-making power of small cultivator womenwas found to be affected by adoption. Beforeadoption, women were hardly consulted by theirmale counterparts while the majority of decisionswere made jointly after adoption. Hence, theintroduction of the technology has helped achievegreater participation and empowerment of thesewomen.

Food and Nutritional Security Impacts

With regard to the effects of microirrigation onfood and nutritional intake, our data reveal thatmainly the small cultivator category of farmershad benefited in terms of access to freshvegetables and improvement in nutritional intake.This was because of the direct involvement ofwomen small cultivators in vegetable cultivationon the homestead farm, which earlier used toremain barren. The adoption of drip and buckethelped these women grow vegetables, which wasused primarily for household consumption, whilethe surplus was sold.

31

As women small cultivators were involveddirectly in market transactions, they had accessto and control over the revenue generated fromthe sale of vegetables. This access and controlover income had a positive impact on overallhousehold food security, as a majority (82percent) of these women spent most of therevenue on household food items. For landlessfamilies, the intervention of drip and bucket didnot have any effect on their normal food intakehabits. Likewise, large cultivator householdspreferred to buy their vegetables as they used to

do earlier, hence no change has been found intheir eating habits.

This implies that the effect of the adoption onimprovement in household food security andnutritional intake is more pronounced for thosefarmers who adopt bucket and drip for homesteadvegetable cultivation. This finding corroboratesthe observation of Upadhyay and Samad (2004)who in their study of microirrigation in the westernhills of Nepal noted that women farmers wereable to ensure improvements in their vegetableand fruit intakes after adopting drum kits.

Conclusions and Implications

Economics of microirrigation systems

The study indicates that, for all the cropsconsidered, on average, the use of microirrigationtechnologies resulted in a significant productivityimprovement (economic gain) over the traditionalmethods of irrigation. Moreover, the yieldresponse to the conventional drip systems (forbanana and cotton) was significantly superior tothat of low-cost drip systems, implying that thelow-cost microirrigation technologies cannot beregarded as ends in themselves but as steppingstones for adopting the conventional systems,which are technically robust and economicallymore rewarding.

Determinants of microirrigation adoption

Technical and economic efficiency is only one ofthe many variables that influence microirrigationadoption decisions of farmers. If technical andeconomic efficiency alone is the maindeterminant of adoption, microirrigationtechnologies would have dominated the traditionalirrigation methods. The successful adoption ofmicroirrigation requires, in addition to technicaland economic efficiency, two additionalpreconditions:

1. The target beneficiaries need to be aware orknowledgeable about the technical andeconomic superiority of the technologies. Thismay be achieved through extension servicesin the form of demonstrations, workshops,etc. Farmers’ own attributes such as level ofeducation may also augment or complementthe public extension services, as educatedfarmers are active information seekers andexperimenters.

2. The technologies need to be accessible tothe potential users. Awareness or knowledgedoes not guarantee actual adoption unlessthe technologies are made accessible to thefarmers through institutional support systems.

The most important variables influencingmicroirrigation adoption decisions in the presentcontext are:

1. Years of schooling of the household head. Asthe level of education of the household headincreases the likelihood of adoptingmicroirrigation technologies increases. Thisconfirms the fact that microirrigationtechnologies need special technical andmanagerial skills for proper utilization. Giventhe fact that the poorer section of the farming

32

population tends to be less educated, specialtraining programs need to be instituted toenable poor people adopt and successfullymanage microirrigation systems.

2. Access to groundwater. As expected,ownership of dug wells and/or bore wellssignificantly increases the probability ofadoption of microirrigation technologies.Moreover, as the depth of dug wells and thehorsepower of pumps owned by farmersincrease the likelihood of adoptingmicroirrigation technologies increases. Thedepth of well and horsepower of pumpindicate the degree of water scarcity and theenergy-saving motives of the farmer.

3. Cropping pattern. The study indicates that thehigher the share of cereals and pulses in thecropping pattern the lower the probability ofadopting microirrigation technologies. Thisimplies that farmers cultivating staple foodcrops are largely excluded from the benefitsof innovations in microirrigation technologies.In the long run, this scenario may bereversed through two possible developments.First, microirrigation engineers may innovatesystems better suited for staple food crops,which currently do not benefit much frommicroirrigation applications. Second, thefarmers may shift their cropping pattern inorder to benefit from microirrigationtechnologies. The latter response will havesignificant implications on the waterresources economy of a watershed or aregion.

4. Additional sources of income. As the share ofincome from other sources than farming(excluding wage-labor income) increases, theprobability of adopting low-cost microirrigationtechnologies increases. This shows theimportance of cash in the initial adoptiondecisions of farmers.

5. Social and poverty status. Despite thetechnical transformation of microirrigationtechnologies to make them pro-poor, well-to-do farmers still have a significantly higherprobability of adopting microirrigation

technologies. In addition, farmers belonging tothe high-caste category have more chance ofadopting microirrigation technologies.

Microirrigation technologies, poverty andwomen

The largest proportion of microirrigationadopters belongs to the relatively very richgroup of farmers. This is especially so inMaharashtra. In Gujarat the situation is a bitmilder, i.e., the adoption is not confined to therichest group but extends to the middle andrich farmers. However, in both locations, thepoor and the poorest sections of the farmingpopulation have not benefited much frominnovations in microirrigation technology. Oneview widely held among development NGOs inIndia and elsewhere is that the technicaltransformation of the conventional, capital-intensive systems into low-cost alternativesgreatly improves the accessibility of thesetechnologies to the poor. Even though this claimis theoretically or potentially true, the observedadoption pattern in the study locations does notlend full support to this view. Though there isslight improvement, at the moment the majority ofthe adopters of low-cost microirrigationtechnologies are those farmers belonging to themiddle, rich and the richest groups.

Thus, reducing the cost alone is not enoughto improve the poverty outreach of microirrigationtechnologies. Three factors limited poor farmer’saccess to low-cost microirrigation technologies.First, the available low-cost microirrigationsystems are suited to crops that are not popularamong poor farmers. Poor farmers tend to allot asignificant proportion of their area to staple foodcrops such as cereals and pulses. Second, theirsocioeconomic attributes (e.g., low level ofeducation, being a member of a low-caste group,low poverty status, etc.) limit their access toinformation, ultimately hindering their access tomicroirrigation technologies. Last, limited accessto resources, specifically to groundwater,quantitatively and qualitatively, hinders poorfarmers from successfully adopting low-costmicroirrigation technologies.

33

Analysis of impacts of microirrigation onwomen in both study regions revealed thatwomen of small cultivator households benefitedmost in terms of access to and control overincome and household nutritional security.Landless, wage earner women experiencednegative impacts of microirrigation technology astheir earning opportunities as wage earners hadsignificantly declined because of reduction inweeding requirements. Women of large cultivatorhouseholds benefited since their workload asfamily labor, particularly in weeding, declined.

Microirrigation use and sustainable utilizationof groundwater resources

The sustainability impact of microirrigation usedepends on:

1. the magnitude of the overall productivity gainfollowing the shift from traditional methods ofirrigation to microirrigation systems (or,indirectly, the volume of water that is savedfor other irrigation uses due to improvementsin the productivity of water),

2. the behavior of adopters following the shift orthe pattern of use of the saved water, and

3. the type and potential number of adopters.

The study indicates that the use ofmicroirrigation technologies increases themarginal productivity of water. The improvementin the marginal productivity of water coupled withthe effect of subsidy schemes that indirectly playthe role of reducing the marginal cost or price ofwater further increases the demand for irrigationwater. Thus, a rational farmer continues toemploy more water in the agricultural production

process until the Value of Marginal Product ofwater equals the price of water.

Specifically, the farmers are expected torespond in one or the other of the following waysdepending on their prevailing circumstances.First, those already suffering from frequent cropfailure or yield losses due to water shortage willmake use of the saved water to obtain a normalharvest or to minimize yield losses. Second,those irrigating only part of their potentiallyirrigable fields due to the inadequacy of water willmake use of the saved water to increase theirrigated area and hence reap more economicgain. Third, the high marginal productivity of watermay spark the demand for more groundwaterresources thereby increasing investments ingroundwater development. The second and thethird scenarios may lead to groundwater overdraft,a fact reported by most of the sample farmers inMaharashtra. They claim that despite 15 years ofexperience with microirrigation, the groundwaterlevel has substantially declined and theconcentrations of wells have increased.

The other remarkable impacts, with significantimplications for the sustainable use ofgroundwater resources, observed in the studylocations following the adoption of microirrigationtechnologies are changes in cropping pattern,cropping intensity and/or crop diversity.Microirrigation adoption led to the prominence ofhigh-value, water-intensive crops in the farmingsystem, and further improved the croppingintensity by enabling the production of crops inthe summer or Rabi. Thus, in the long run, thesustainability objective may conflict with thepoverty reduction and food security objectivesunless a proper regulatory mechanism isinstituted.

34

35

Literature Cited

Aldrich, J. H.; and Nelson, F. D. 1984. Linear probability, logit and probit models. London: Sage Publications.

Amemiya, T. 1981. Qualitative response models: A survey. Journal of Economic Literature 19: 1483-1536.

Bilgi, M. 1999. Socio-economic study of the IDE promoted microirrigation systems in Aurangabad and Bijapur.Prepared for Swiss Development Cooperation, New Delhi, India.

Buttel, F. H.; Larson, O. F.; and Gillespie, G. W. 1990. The sociology of agriculture. New York: Greenwood Press,Inc.

Caswell, M. F. 1999. The adoption of low volume irrigation technologies as a water conservation tool. WaterInternational 14: 19-26.

CIMMYT (International Maize and Wheat Improvement Center). 1988. From agronomic data to farmerrecommendations. An economics training manual. Revised Edition. Mexico City: CIMMYT.

Debertin, D. L. 1986. Agricultural production economics. New York: Macmillan Publishing Company.

Dhawan, B. D. 2002. Technological change in irrigated agriculture: A study of water saving methods. New Delhi:Commonwealth Publishers.

Feder, G.; Just, R. E.; and Zilberman, D. 1985. Adoption of agricultural innovations in developing countries: Asurvey. Economic Development and Cultural Change 33, 255-298.

Greene, W. H. 2000. Econometeric analyses. New York: Macmillan.

Henry, C.; Lapenu, C.; Zeller, M.; and Sharma, M. 2003. Microfinance poverty assessment tool. Washington, D.C.:The World Bank.

Kennedy, P. 2001. A guide to econometrics. Fourth Edition. Cambridge, Massachusetts: The MIT Press.

Leagans, P. J. 1979. Adoption of modern agricultural technology by small farm operators: An interdisciplinarymodel for researchers and strategy builders. Ithaca, New York: Cornell University.

Molden, D.; Hammond, M-R.; Sakthivadivel, R.; and Makin, I. 2003. A water productivity framework forunderstanding and action. In Water productivity in agriculture: Limits and opportunities for improvement, eds.J. W. Kijne, R. Barker and D. Molden. Wallingford, Oxon, UK: CABI Publishing.

Namara, R. E.; Weligamage, P.; and Barker, R. 2003. Prospects for adopting system of rice intensification in SriLanka: A socioeconomic assessment. Research Report 75. Colombo, Sri Lanka: International WaterManagement Institute.

Narayanamoorthy, A. 1997. Economic analyses of drip irrigation: An empirical analysis from Maharashtra. IndianJournal of Agricultural Economics 52 (4): 728-739.

Narayanamoorthy, A. 2003. Averting water crises by drip method of irrigation: A study of two water intensive crops.Indian Journal of Agricultural Economics 58 (3): 427-437.

Neil, S. P.; and Lee, D. R. 2001. Explaining the adoption and disadoption of sustainable agriculture: The caseof cover crops in northern Honduras. Economic Development and Cultural Change 49 (4): 793-820.

Polak, P.; Nanes, B.; and Adhikari, D. 1997. A low cost drip irrigation system for small farmers in developingcountries. Journal of the American Water Resources Association 33 (1): 119-124.

Rogers, E. M. 1995. Diffusion of innovations. Fourth Edition. New York: The Free Press, a division of Simon andSchusfer, Inc.

Sakks, R. K. 2001. Predicting drip irrigation use and adoption in a desert region. Agricultural Water Management51 (2001): 125-142.

36

Shah, T.; and Keller, J. 2002. Micro-irrigation and the poor: A marketing challenge in smallholder irrigationdevelopment. In Private irrigation in sub-Saharan Africa: Regional seminar on private sector participation andirrigation expansion in sub-Saharan Africa, 22-26 October 2001, Accra, Ghana. Proceedings, eds. Hilmy Sallyand Charles L. Abernethy. Colombo, Sri Lanka: International Water Management Institute, Food and AgricultureOrganization of the United Nations, and ACP-EU Technical Centre for Agricultural and Rural Cooperation.

Shrestha, R. B.; and Gopalakrishnan, C. 1993. Adoption and diffusion of drip irrigation technology: Aneconometric analysis. Economic Development and Cultural Change 41 (2): 407-418.

Sivanappan, R. K. 1994. Prospects of microirrigation in India. Irrigation and Drainage Systems 8 (1): 49-58.

Upadhyay, B.; Samad, M. 2004. Livelihoods and gender roles in drip-irrigation technology: A case of Nepal.Working Paper 87. Colombo, Sri Lanka: International Water Management Institute.

Upadhyay, B. 2004. Gender aspects of smallholder irrigation technology: Insights from Nepal. Journal of AppliedIrrigation Science Vol. 39 (2). Department of Rural Engineering and Natural Resources Protection and the DLG-Varrlags-GMBH.

Upadhyay, B. 2003. Drip irrigation: An appropriate technology for women. Appropriate Technology Vol. 30 (4).

Verma, S.; Tsephal, S.; and Jose, T. 2004. Pepsee systems: Grass root innovation under groundwater stress.

Water policy 6 (2004): 1-16.

37

Appendix 1

Salient Features of the Study Locations

38

TAB

LE A

1.N

ames

of

villa

ges

and

talu

kas

in t

he s

tudy

dis

tric

ts.

Adm

inis

trat

ive

unit

Nam

e of

the

stu

dy

dist

rict,

talu

ka o

r vi

llage

Sta

teM

ahar

asht

raG

ujar

at

Dis

tric

tJa

lgao

nR

ajko

tJu

naga

dh

Talu

kaC

hopd

aJa

lgao

nJa

mne

rP

acho

raR

ave

rJe

tpur

Kot

ada

San

gani

Raj

kot

Mal

iya

Vill

age

Dev

agon

Edg

aon

Gon

dega

onK

alam

sare

Viv

are

Um

rali

Ram

ond

Kho

khad

adad

Bab

ra

Dha

nora

Dex

ieS

heng

ole

Loha

tar

Ker

hale

Loth

ada

Kuk

asw

ada

Par

gaon

Vid

gaon

Loha

reLa

duli

Loha

riM

oti

Dha

nej

Nan

i D

hane

j

Sha

nti

Pur

a

39

TAB

LE A

2.B

asic

info

rmat

ion

on d

emog

raph

ics,

land

use

and

irrig

atio

n in

fras

truc

ture

in t

he s

tudy

loca

tions

.

Cha

ract

eris

ticR

ajko

tJu

naga

dhG

ujar

atJa

lgao

nM

ahar

asht

ra

1992

2001

1992

2001

1992

2001

1992

2001

1992

2001

Rur

al p

opul

atio

n1,

330,

156

1,56

2,38

31,

615,

483

1,73

7,10

127

,063

,521

31,7

40,7

672,

312,

965

2,62

7,31

548

,395

,601

55,7

77,6

47

Urb

an p

opul

atio

n1,

183,

966

1,59

5,29

377

9,37

671

1,32

614

,246

,061

18,9

30,2

5087

4,66

91,

052,

621

30,5

41,5

8641

,100

,980

Tota

l pop

ulat

ion

2,51

4,12

23,

157,

676

2,39

4,85

92,

448,

427

41,3

09,5

8250

,671

,017

3,18

7,63

43,

679,

936

78,9

37,1

879,

6878

,627

Tota

l geo

grap

hic

area

(ha)

1,12

0,30

01,

120,

300

1,06

0,70

01,

060,

700

196,

11,7

0019

,611

,700

1,16

3,90

01,

163,

900

30,7

58,3

0030

,758

,300

Net

are

a so

wn

(ha)

723,

700

737,

400

573,

000

587,

200

9,34

7,10

09,

667,

400

830,

200

852,

500

17,8

94,8

0017

,636

,400

Are

a so

wn

mor

e th

an o

nce

(ha)

49,8

0070

,700

60,2

0084

,000

1,21

0,60

01,

476,

400

140,

100

480,

600

2,23

8,60

04,

619,

400

Gro

ss c

ropp

ed a

rea

(ha)

773,

500

808,

100

633,

200

671,

200

10,5

57,7

0011

,143

,800

970,

300

1,33

3,10

020

,133

,400

22,2

55,8

00

Net

irrig

ated

are

a: S

urfa

ce ir

rigat

ion

(ha)

13,4

0040

,900

7,50

012

,200

507,

500

651,

900

7,40

011

,000

980,

400

1,05

0,20

0

Net

irrig

ated

are

a: W

ell i

rrig

atio

n (h

a)15

1,60

010

7,60

010

2,50

014

0,20

01,

930,

100

2,43

0,50

013

1,60

014

4,40

01,

745,

500

2,09

0,00

0

Tota

l net

irrig

ated

are

a (h

a)16

5,00

014

8,50

011

0,00

015

2,40

02,

437,

600

3,08

2,40

013

9,00

015

5,40

02,

725,

900

3,14

0,20

0

Are

a irr

igat

ed m

ore

than

onc

e (h

a)5,

600

63,0

0012

,300

9,40

047

2,90

075

8,20

029

,600

16,7

0053

8,90

055

3,00

0

Gro

ss a

rea

of ir

rigat

ed c

rops

(ha

)17

0,60

021

1,50

012

2,30

016

1,80

02,

910,

500

3,84

0,60

016

8,60

017

2,10

03,

264,

800

3,69

3,20

0

Gro

undw

ater

stru

ctur

es (

1994

)a-

-72

2,95

8-

1,35

0,27

4

Gov

ernm

ent t

ube

wel

ls (

1998

)-

--

169

35,1

49

Priv

ate

tube

wel

ls (

1998

)-

--

1,70

536

,143

Oth

er g

over

nmen

t wel

ls (

1998

)-

--

039

,746

Oth

er p

rivat

e w

ells

(19

98)

--

-62

,121

1,23

7,06

7

Oil

engi

ne p

ump

sets

(199

8)77

512

2,42

2

Ele

ctric

pum

p se

ts (

1998

)48

,531

1,08

0,95

4

Dis

trict

s w

ith r

ecor

ded

decl

ine

in w

ater

leve

l (19

99-2

002)

Yesb

Yesb

5Ye

sb27

Ran

ge o

f pre

-mon

soon

gro

undw

ater

decl

ine

in m

eter

s (1

999-

2002

)-

0.45

-3.3

20.

15-2

.68

Num

ber o

f wel

ls d

rille

d as

of N

ovem

ber 2

000

--

579

693

a Inc

lud

es g

roun

dw

ater

dev

elop

ed f

or o

ther

pur

pos

es t

han

irrig

atio

n.b I

ndic

ates

whe

ther

the

stu

dy

loca

tion

was

inc

lud

ed i

n th

e lis

t of

dis

tric

ts w

ith r

ecor

ded

dec

line

in w

ater

lev

el (

1999

-200

2).

40

TAB

LE A

3.C

ropp

ing

patte

rns

of t

he s

tudy

loca

tions

fo

r th

e ye

ar 1

991/

1992

.

Raj

kot

Juna

gadh

Guj

arat

Jalg

aon

Mah

aras

htra

Cat

egor

yh

a%

ha

%h

a%

ha

%h

a%

Cer

eals

105,

000

13.6

91,0

0014

.43,

731,

900

35.4

366,

500

37.8

9,99

0,60

049

.6

Pul

ses

11,6

001.

59,

700

1.5

882,

700

8.4

174,

900

18.0

3,08

0,80

015

.3

Sug

arca

ne4,

500

0.6

9,50

01.

517

2,60

01.

620

,400

2.1

588,

600

2.9

Con

dim

ents

and

spi

ces

11,5

001.

54,

900

0.8

139,

200

1.3

4,40

00.

515

0,70

00.

7

Ban

ana

00.

040

00.

0632

,300

0.3

44,8

004.

672

,400

0.4

Oth

er f

ruit

crop

s80

00.

16,

000

1.0

67,5

000.

62,

400

0.3

171,

500

0.9

Veg

etab

les

7,00

00.

93,

400

0.5

105,

700

1.0

3,50

00.

423

8,50

01.

2

Gro

undn

ut43

4,30

056

.141

8,30

066

.11,

976,

100

18.7

53,6

005.

573

9,00

03.

7

Oth

er o

il se

eds

36,7

004.

715

,000

2.4

888,

400

8.4

93,5

009.

61,

534,

600

7.6

Cot

ton

90,3

0011

.731

,400

4.9

1,16

4,30

011

.020

4,00

021

.02,

759,

100

13.7

Oth

er f

iber

cro

ps0

0.0

00.

08,

800

0.08

200

0.02

38,7

000.

2

Nar

cotic

s an

d pl

anta

tion

crop

s70

00.

110

00.

0117

9,00

01.

70

0.0

8,70

00.

04

Fod

der

crop

s71

,100

9.2

43,3

006.

81,

205,

500

11.4

2,00

00.

275

1,50

03.

7

Oth

ers

00.

020

00.

033,

700

0.04

100

0.01

7,70

00.

04

Tota

l77

3,50

010

063

3,20

010

010

,557

,700

100

970,

300

100

20,1

32,4

0010

0

41

Appendix 2

Derivation of the MPP from the Fitted Transcendental Production Function

Banana

( )SxWPExSNxKLCMICNMIPKNPESWb ePKNPESAWY 1110987654321654321 γγγγγγγγγγγαααααα ++++++++++=

bbW YSWwyMPP )/(/ 1111 γγα ++=∂∂=

( ) bbS YWPESSyMPP 111022 // γγγα +++=∂∂=

( ) bbPE YSPEPEyMPP 1033 // γγα ++=∂∂=

( ) bbN YKNNyMPP 944 // γγα ++=∂∂=

( ) bbK YNKKyMPP 955 // γγα ++=∂∂=

( ) bbP YPPyMPP 66 // γα +=∂∂=

bbCNMI YCNMIyMPP 7/ γ=∂∂=

bbLCMI YLCMIyMPP 8/ γ=∂∂=Where:

Yb = Banana yield per ha; A= Constant term; W = Weeding labor input (man-days per ha); S = Seed cost (Rs

per ha); PE = Pesticide (l per ha); N = Nitrogen (kg per ha); K = Potassium (kg per ha); P = Phosphorus (kg per

ha); MPPw

= Marginal Physical Product of weeding labor input; MPPs = Marginal Physical Product of seed

input; MPPPE

= Marginal Physical Product of pesticide input; MPPN

= Marginal Physical Product of nitrogen;MPP

K = Marginal Physical Product of potassium; MPP

P = Marginal Physical Product of phosphorus; MPP

CNMI =

Marginal Physical Product of conventional microirrigation system; MPPLCMI

= Marginal Physical Product of low-

cost microirrigation system.

Groundnut

( )wXsdripmispCNSPPUNPESWGN ePUKNPESAWY 987654321654321 γγγγγγγγγαααααα ++++++++=

( ) GNGNW YSWWyMPP 911 // γγα ++=∂∂=

( ) GNGnS YWSSyMPP 922 // γγα ++=∂∂=

( ) GNGNPE YPEPEyMPP 33 // γα +=∂∂=

( ) GNGNN YNNyMPP 44 // γα +=∂∂=

GNGNK YKKyMPP )/(/ 5α=∂∂=

42

( ) GNGNPU YPUPUyMPP 56 // γα +=∂∂=

GNGNCNSP YCNSPyMPP 6/ γ=∂∂=

GNGNMISP YMISPyMPP 7/ γ=∂∂=

GNGNDRIP YDRIPyMPP 8/ γ=∂∂=

Where:

YGN

= Groundnut yield per ha; A = Constant term; W = Weeding labor input (man-days per ha); S = Seed cost(Rs per ha); PE = Pesticide (l per ha); N = Nitrogen (kg per ha); K = Potassium (kg per ha); PU = Pumping

hours (hour per ha); MPPw

= Marginal Physical Product of weeding labor input; MPPs = Marginal Physical

Product of seed input; MPPPE

= Marginal Physical Product of pesticide input; MPPN

= Marginal PhysicalProduct of nitrogen; MPP

K = Marginal Physical Product of potassium; MPP

PU = Marginal Physical Product of

pumping hours; MPPCNSP

= Marginal Physical Product of conventional sprinkler system; MPPMISP

= Marginal

Physical Product of microsprinkler system; MPPDRIP

= Marginal Physical Product of drip system.

Cotton

( )PUSWPESPEWSLCMIMITDCNDPPUPNPESWCN ePUPNPESAWY 13121110987654321654321 γγγγγγγγγγγγγαααααα ++++++++++++=

( ) CNCNW YPESWWyMPP 121011 // γγγα +++=∂∂=

( ) CNCNS YPUPEWSSyMPP 13111022 // γγγγα ++++=∂∂=

( ) CNCNPE YWSPEPEyMPP 121133 // γγγα +++=∂∂=

( ) CNCNN YNNyMPP 44 // γα +=∂∂=

( ) CNCNP YPPyMPP 55 // γα +=∂∂=

( ) CNCNPU YSPUPUyMPP 1366 // γγα ++=∂∂=

CNCNCNDP YCNDPyMPP 7/ γ=∂∂=

CNCNMITD YMITDyMPP 8/ γ=∂∂=

CNCNLCMI YLCMIyMPP 9/ γ=∂∂=Where:

YCN

= Cotton yield per ha; A = Constant term; W = Weeding labor input (man-days per ha); S = Seed cost (Rs

per ha); PE = Pesticide (l per ha); N = Nitrogen (kg per ha); P = Phosphorus (kg per ha); PU = Pumping hours(hour per ha); MPP

w = Marginal Physical Product of weeding labor input; MPP

s = Marginal Physical Product of

seed input; MPPPE

= Marginal Physical Product of pesticide input; MPPN

= Marginal Physical Product of

nitrogen; MPPP

= Marginal Physical Product of phosphorus; MPPPU

= Marginal Physical Product of pumpinghours; MPP

CNDP = Marginal Physical Product of conventional drip system; MPP

MITD = Marginal Physical

Product of microtube drip system; MPPLCMI

= Marginal Physical Product of low-cost microirrigation system.

80. Robbing Yadullah’s Water to Irrigate Saeid’s Garden: Hydrology and Water Rightsin a Village of Central Iran. François Molle, Alireza Mamanpoush and MokhtarMiranzadeh. 2004.

81. Inadequacies in the Water Reforms in the Kyrgyz Republic: An InstitutionalAnalysis. Mehmood Ul Hassan, Ralf Starkloff and Nargiza Nizamedinkhodjaeva.2004.

82. Valuing Nutrients in Soil and Water: Concepts and Techniques with Examplesfrom IWMI Studies. Pay Drechsel, Mark Giordano and Lucy Gyiele. 2004.

83. Spatial Variation in Water Supply and Demand Across River Basins of India.Upali A. Amarasinghe, Bharat R. Sharma, Noel Aloysius, Christopher Scott,Vladimir Smakhtin and Charlotte de Fraiture. 2004.

84. An Assessment of Small-scale Users’ Inclusion in Large-scale Water UserAssociations of South Africa. Nicolas Faysse. 2004.

85. The Use of Remote-Sensing Data for Drought Assessment and Monitoring inSouthwest Asia. P. S. Thenkabail, M. S. D. N. Gamage and V. U. Smakhtin, 2004.

86. Strategies for the Management of Conjuctive use of Surface Water andGroundwater Resources in Semi-arid Areas: A Case Study from Pakistan. AsadSarwar Qureshi, Hugh Turral and Ilyas Masih. 2004.

87. Economics and Politics of Water Resources Development: Uda Walawe IrrigationProject, Sri Lanka. François Molle and Mary Renwick. 2005.

88. "Bright Spots" in Uzbekistan, Reversing Land and Water Degradation WhileImproving Livelihoods: Key Developments and Sustaining Ingredients for TransitionEconomies of the former Soviet Union. Andrew Noble, Mohammed ul Hassan andJusipbek Kazbekov. 2005.

89. Planning for Environmental Water Allocations: An Example of Hydrology-basedAssessment in the East Rapti River, Nepal. V. U. Smakhtin and R. L. Shilpakar.2005.

90. Working Wetlands: Classifying Wetland Potential for Agriculture. Matthew P.McCartney, Mutsa Masiyandima and Helen A. Houghton-Carr. 2005.

91. When “Conservation” Leads to Land Degradation: Lessons from Ban Lak Sip,Laos. Guillaume Lestrelin, Mark Giordano and Bounmy Keohavong. 2005.

92. How Pro-Poor are Participatory Watershed Management Projects?—An Indian CaseStudy. Mathew Kurian and Ton Dietz. 2005.

93. Adoption and Impacts of Microirrigation Technologies: Empirical Results fromSelected Localities of Maharashtra and Gujarat States of India. Regassa E. Namara,Bhawana Upadhyay and R. K. Nagar. 2005.

Research Reports

Regassa E. Namara, Bhawana Upadhyay and R. K. Nagar

IWMI is a Future Harvest Centersupported by the CGIAR

Adoption and Impacts ofMicroirrigation TechnologiesEmpirical Results from SelectedLocalities of Maharashtra andGujarat States of India

93

RESEARCHR E P O R T

SM

IWMI is a Future Harvest Centersupported by the CGIAR

Postal Address:P O Box 2075ColomboSri Lanka

Location:127, Sunil MawathaPelawattaBattaramullaSri Lanka

Tel:+94-11-2787404

Fax:+94-11-2786854

E-mail:iwmi@cgiar.org

Website:http://www.iwmi.org

I n t e r n a t i o n a lWater ManagementI n s t i t u t e

ISSN 1026-0862ISBN 92-9090-602-2

I n t e r n a t i o n a lWater ManagementI n s t i t u t e