A management perspective on the performance of the irrigation subsector.

INVESTMENT IN IRRIGATION has been immense in the past. Estimated average annual investments of US$ 15 billion makes irrigation the largest subsector of the agricultural sector, that is itself by far the largest sector of development investment. Since the mid-1960s the awareness spread that the performance of irrigation investments was far below its potential. The size of this underperformance is well represented by Seckler's alarming conclusion that the average irrigation investment costs twice as much, and delivers no more than half the benefits specified in the plans.<strong>THE PROBLEM DEFINITION</strong>Simultaneously with the increased awareness about underutilization, the awareness increased that the level of management of the systems was backward compared to the construction efforts and expertise. The underutilization was considered not only a technical, but also a managerial problem. Essentially three pilot studies in the late 1970s in the Philippines, Sri Lanka and India have provided the few available data to proof such potential for performance improvement through improved management. Yet, this potential for a sustainable "water revolution" remains to date largely as it was, because the evidence of these three experiments was not repeated nor sustained.From the perspective of many engineers, the management issue in irrigation has remained therefore, to a large extent, imaginary. There has remained thus a serious disjuncture in the perspectives of many irrigation professionals. Many of them have argued for the need for a more objective perspective on irrigation's performance to reunite the different professional perspectives, and as a prerequisite for the identification of relevant improvements. The topic of this study is such an improved insight in the management of irrigation, and ways to improve its performance.<strong>THE OBJECTIVES</strong>In addressing these issues, this study adopts the following two objectives: 1) the identification of generalized directions of management change for performance improvement in the irrigation subsector; and 2) the testing of an analytical framework for irrigation management.Addressing these objectives requires firstly an effort to fill the fore mentioned gap toward the concept of irrigation management. Therefore, the concepts of management and control processes and conditions of an existing analytical management framework are translated for irrigation. Together they form this paper's so-called management perspective. Subsequently, this analytical framework is applied to irrigation.<strong>EXISTING IRRIGATION MANAGEMENT CONCEPTS</strong>Few explicit efforts to develop irrigation management concepts appear to exist. Most concepts focus on the formal appearance of the organization, its structure. Of the reviewed concepts, only Diemer's approach was a process-oriented approach. All concepts remained vague about the relation between process and structure. None of them tried to take a management perspective, i.e., to consider all relevant factors for irrigation managers. This study's potential contribution is to fill these gaps by taking an explicit management perspective, and by systematically analyzing the relation between process and structure. Besides, other management conditions than structure only are considered such as financial control systems, human resources, and the provision of information and knowledge.<strong>AN INTEGRAL MANAGEMENT PERSPECTIVE</strong>This study's management perspective is based on an integral management framework developed by Kampfraath and his colleagues of the Department of Management Studies of the Wageningen Agricultural University, The Netherlands.The Figure below is a graphic representation of the different steps of this processbased management analysis. The identification of key decisions in regard to water delivery is the first step in the development of this management perspective on irrigation (step I of the Figure below). For irrigation agencies, the management of water is considered the primary irrigation activity and measure of performance evaluation. Therefore, to evaluate the internal management processes in any irrigation system, the relevant key decisions for irrigation have to relate to the water delivery.During the capacity utilization , the seasonal allocation plan, in-seasonal allocation, and the flow regulation are considered to be such key decisions. For the capacity creation , the desired investment objectives, feasible investment objectives, and the functional requirements for the investment were taken as the most relevant key decisions.<img src="/wda/abstracts/i1605_1.gif" height="262" width="600"/>After the definition of the relevant key decisions, the contribution to the overall performance must be established for each of them (step 2). If this contribution is deemed unsatisfactory, the processes leading to the final decisions are analyzed, and the bottlenecks in these processes are identified. The establishment of the so-called levels of sophistication of the key decisions is part of this analysis (step 3). Based on an analysis of the interaction between the processes and the management conditions, those changes in the management conditions are derived that are likely to lead to improved processes. Apart from the organizational structure and rules, this framework also considers such other management conditions as the human resources, their motivation and incentives, the provision of information and knowledge, and the financial control systems. This leads to an identification of the changes needed in management conditions that are likely to result in improved processes, improved decisions, and improved performance (step 4). The last step is then the identification of the required management-control processes to achieve these required improvements in processes and management conditions (step 5).This analytical framework thus links performance, physical processes, decisionmaking processes, management conditions and management control in an analytical sequence. Thus providing an integral "management perspective" on irrigation performance.<strong>DATA COLLECTION</strong>This study's data collection occured during in-depth organizational analyses of two Sri Lankan irrigation organizations, and during comparative studies in Morocco, Sudan, and the Philippines. Besides, less intensive observations were done in India, Malaysia and Pakistan.The data collection on decision-making processes consisted of the interviewing of decision makers in irrigation and other line agencies, ministries, funding agencies and consultant companies. Also reports, files, records and other documentation were reviewed. In addition, a literature survey was done to shape and compare the findings.The following sections give short summaries of the most significant findings and recommendations for the management of the capacity utilization and the capacity creation of irrigation in LDCs. <strong></strong><strong>RESULTS: CAPACITY UTILIZATION</strong>The assessment of the available water supply in the observed irrigation systems tended to occur in an approximate rather than a precise way. They tended to be on the "safe" side -- preferably at a 100 per cent probability, i.e., at no risk--to minimize cultivation risks, and to minimize the related conflicts with the farmers and politicians. This practice preemped the inclusion of the trade-offs between lower risks for the few lucky farmers, and higher risks for more farmers. Other interested parties than the irrigation agency or officer were usually not aware of the exact probabilities of the availability of the water supply. Thus, they did not share the responsibility for any related risks.Contrary to common belief, the assessment of the demand, the allocation of water, and the regulation tended to be demand-driven in all case studies. This decision making was left almost completely to the field level staff. Higher level agency staff made water schedules based on theoretical calculations. These excluded important aspects such as the scarcity of water and the required management inputs by agency staff and farmers to achieve high water efficiencies. These schedules seldom had any value for the actual implementation of water allocation and regulation.Higher level staff thereby tried to minimize their management inputs. Only when complaints occured did staff get involved. To minimize complaints they tended to allow field staff to satisfy the demand for water, and allowed a related superfluous water discharge in all canals (if the supply was available). The main canals thus often transported the maximum discharge. Systematic monitoring and evaluation was not done in any of the case studies, as the extra water in the canals and the "delegation" to field staff did not need any.This minimal management approach of the agencies favored the farmers at head-end reaches along the canals. So the tail enders often had problems in obtaining sufficient water as the design and actual discharge capacities of the canals were insufficient for this type of surplus water allocation.The flow regulation along the main canals appeared an "adhocracy". The individual operators had no contingency instructions on procedures for gate settings relative to the timing and size of flow fluctuations. They operated therefore often by trial-and-error and tended to favor thereby the distribution to the service area under their responsibility. This was done again to satisfy farmers and to minimize complaints, at the expense of the conveyance to downstrean canal reaches. Also gate operators that were responsible for tail-end reaches could not correct systematically such favoring by upstream reaches. The easiest way out for them was to request an increase of the total discharge in the main canal.Getting more water to the end of the canal was then only possible by the allocation of more discharge to the overall system, or, if that were impossible, by rotation or staggering. Introduction of the latter measures required increased management inputs by higher level staff, and occurred only if (a portion of the) farmers, superiors or politicians complained.The above processes were mainly caused by the low motivation of agency staff involved in the capacity utilization. The incentives to be involved appeared to be mainly the following "negative" ones: farmers were never satisfied, a lack of performance-related financial or career incentives, the continuous risk of political interference, and professional and financial incentives for construction and maintenance rather than for the capacity utilization. Similarly, irrigation agencies as a whole had no performance-related incentives, other than the fore mentioned negative ones.The above practices, motivation and incentive constraints were more true in some countries than in others. In Morrocco, the management practices were at a more elevated level compared to the other case studies and some of the above generalizations did not apply to Morocco. Performance-related motivation and incentives were observed to be somewhat higher because of a more sophisticated management, with individual billing, and a volumetric water delivery to farmers. Yet, also in Morocco, the agencies appeared to have no financial incentives to manage the allocation and regulation along the main system, in order to prevent obvious and known water losses.Main recommendations. In all case studies, improvement of the capacity utilization would require increased inputs by higher level agency staff. This would require that they, as well as field staff, become more motivated for this type of work. This seemed unlikely to occur if the agencies themselves would not become more interested in and accountable for the water-delivery performance.The overall recommended directions for institutional reform to improve the capacity utilization were the following:1. A decentralization of the irrigation agencies. This would allow greater information exchanges at lower levels between farmers and agency, and between different agency levels;2. More financial dependence of the irrigation agency on the water-delivery performance. This would introduce some accountability for the waterdelivery performance. For example through an increased dependence on service payments by farmers. More financial independence of the irrigation agencies means also a decreased dependence on the judiciary budgetary allocation by the government;3. A more performance-oriented human resources management, such as performance-based incentive systems and career development, especially for higher level staff. This would require a decentralization of the related authority to the agency;4. A more explicit and specific mission statement;5. External public monitoring for more systematic accountability (if no financial or other accountability to the clients exists);6. More transparency of and thus accountability for the performance of the regulation through a separate, central "regulation unit";7. If WUGs are to be functional, they need a more powerful position in the water- related decision-making processes than currently observed in all case studies. This could be achieved either through more administrative authority, or through financial accountability to the WUGs. An ultimate step as the transfer of the ownership of (part of) the system (and possibly the agency) would provide the collectivity of farmers with even stronger powers to make the managing agency accountable for the performance during the capacity utilization;8. More appropriate government regulations and related enforcement to reduce the observed adverse incentives in some more independent irrigation agencies.Given the involved interests of agencies, their staff and the farmers, the above changes can only be realized if they get serious support from political and donor levels.<strong>RESULTS: CAPACITY CREATION</strong>The desired investment objectives. Decision taking on the desirable investment objectives was often observed to be done single-handedly by national politicians. Usually, donor. staff in consultation with consultants and agency staff prepared such decisions. This preparation usually left little time and room for participatory interactions with other interest groups.Politicians often determined such politically relevant objectives as the site identification and the selection of beneficiaries. The political pressure thereby caused the professional guidance to become sometimes ineffective.The acquisition of external funding was observed to be the prevalent political and agency priority. It dominated the other desired objectives, other than those of political importance. Because of this priority for external funding, the funding agency had, in principle, and in practice, a large influence on the determination of the desirability of the investment objectives. Thus, the desirability of such investment objectives as the project size and the performance of the water delivery and agricultural production was in all case studies largely at the discretion of the donor staff or consultants.As a result, also the interests of farmers and other local interests were unlikely, and were observed not, to be adequately represented in the decision making on the desirable investment objectives. Often the desirability from the farmers' perspective was considered equal to the maximum funding level as the funds were perceived as "handouts" to localized voters.In combination with an observed supply-driven availability of financial resources for irrigation investment, such politically dominated processes and the related attitudes worked against choices for less capital-intensive, more effective investments in, for example, water management and conservation. This applied to all case studies.Yet, the major gap in this decision making seemed the observed absence of an explicit definition of the desired performance levels for the new investments. The widespread and long-established experiences with ineffective capacity utilization in irrigation made this absence all the more striking. Even if donors, governments or consultants were aware of the unlikeliness of achieving the assumed performance improvements in specific systems or projects, they were observed to ignore such considerations in the investment selection and design process. Assumed performance targets of irrigation investments were kept implicit in all case studies.The likely commitment of such stakeholders as the national politicians, governments and agencies to the implicitly defined performance targets was almost nil (except for the few who were internally motivated). The more so given the lack of incentives to achieve them, while ample incentives were observed to relate to the acquisition of new investments (in an environment with abundant availability of financial resources for irrigation investment). In all case studies, the commitment of staff of the national planning agency, the irrigation ministry and the irrigation agency toward performance improvements was observed to be almost absent. The widespread underutilization of irrigation capacities in the past did not seem to have led to a stronger conditionality toward the quality of subsequent investment decisions. The underlying reason seemed the conflict of the quality of investment decisions with the fund-channeling function of the donor agencies, i.e., with the quantity of the investments.The feasible investment objectives. Financial resources for irrigation seemed abundant mainly because of the nature of the feasibility and appraisal assessments. In all case studies, the feasibility and appraisal assessments were observed to occur after the political decision to undertake the project. The different steps and methodologies served merely to justify the decision. Consideration of the feasibility of alternative types of projects, project sites, or a more phased development to achieve the same objectives, were ignored in all case studies.The preparation of the decision about what was feasible and what not, was observed to be mainly the task of donor staff and consultants. Sometimes because they were considered more "independent" than staff of the recipient country or agency, in other cases because they would prevent likely delays in loan disbursements. Yet, it was observed to be very difficult, if not impossible, for them to determine the true feasibility, especially of the assumed performance improvement. Recipient agency and government staff tended to represent vested interests to realize the funding, and were unlikely to provide any information counteracting these interests. Even in the few observed cases where they were willing to do so, they were usually not asked to. Also the assessment experts and donor staff, who were driven by their organization's targets, were not interested in an absolutely neutral feasibility assessment.The assumed performance improvements and other optimistic assumptions were thus typically not justified. They were kept implicit. Cost-benefit and sensitivity analyses were not allowed in any of the case studies to classify a project as unfeasible. They seemed therefore to have lost their functionality for an objective assessment of investment feasibility and appraisal. Instead, they were used to facilitate subsidies for irrigation investments.The observed funding agencies were observed to have undertaken remarkably little to minimize or counterbalance some of the tendency to be overly optimistic in feasibility assessments. Rather than demanding explicit evidence of assumed performance improvements, virtually the only check and balance mechanism within the development banks was observed to be the mild "peer reviews". These meetings tended to be chaired by persons who were primarily responsible for the quantity of loans, rather than for their quality.Performance targets for investments were implicitly set during feasibility decision making, and tended to be mainly donor-driven. Commitment to, or awareness of, these targets by staff of national governments and agencies was very low to zero.Justifications for why an investment would not become another failure were mostly conceptual, rather than related to real-life. The different conceptual approaches developed to overcome the "management gap" (such as parallel field canals, on-farm water management, O&M manuals, water-management consultants, farmer participation and monitoring and evaluation) did not increase the commitment of the agency and government as they did not touch upon the performance and accountability issues. In fact, these solutions increased the donors' influence in actual investment planning and design, whereby the agencies felt increasingly less responsible, resulting in a diminishing commitment from their side. Sequential conceptual solutions established in a donordriven mode, seemed to have produced progressively less and less commitment to their actual feasibility by the national agency staff.A logical and related effect of the observed manipulation of the assumptions pertaining to the economic internal rate of return (EIRR) was the increased lack of any control over capital expenditures from the national point of view. Limits on expenditures per resource unit were observed to be non-existent in all case studies. For example, the maximum investment per settler, per unit of increased agricultural production, per unit of volume stored or regulated, per job created, per area commanded were seldom determined and were thus de-facto based only on political considerations. This led to investment maximization attitudes by irrigation agencies and politicians--at great economic loss for the overall country in current and future generations.The functi