Irrigation management problems derived by a WUA evaluation at the Kpong Irrigation Scheme (KIS) in Ghana

This study examines management problems of WUA C1 of the Kpong Irrigation Scheme (KIS) in Ghana using the 13 evaluation indicators derived from theories of common-pool resources management proposed by Ostrom and Freeman. The evaluation shows that WUA C1 lacks the indicator of water control ability even though C1 has fine irrigation facilities that enable independent water management within WUAs. The reason is that KIS has been managed by Joint System Management (JSM) in which the KIS Office of Ghana Irrigation Development Authority operates and maintains the main canal, main drainage and main road, while WUA C1 operates and maintains the branch canal and below. However, the KIS Office could not fully operate and maintain the irrigation facilities because of budgetary limitations. In irrigation management transfer, it is common to set JSM that allots the management of main canals to the government (or scheme management entity), and the management of lateral canals and below to WUAs. However, from the perspective of WUA organizational evaluation, it is possible that JSM harms the water control ability of WUAs.

Keywords: Common-pool resources management; Irrigation management transfer (IMT); Joint System Management (JSM); WUA

Many public managed irrigation schemes built after World War II could not attain the expected targets in the operation and maintenance (OM) stage. Due to the deterioration conditions of irrigation facilities, the idea of participatory irrigation management (PIM) which requires involvement of farmers in operation, management, and maintenance of the irrigation system was widely applied since the early 1990s (Ishii and Sato [6]; Sato and Sato [15]; Kulkarni and Tyagi [10]). Together with PIM, irrigation management transfer (IMT) has been promoted by the World Bank (Hatcho and Tsutsui [5]). IMT is a process of shifting a number of irrigation management functions from a public agency to private-sector entities, often the beneficiary farmers group [i.e., Water Users Association (WUA)]. The most common form of IMT involves the shifting of management responsibility from a centralized government irrigation agency to a WUA (Svendsen et al. [18]). IMT has been utilized as a tool for irrigation sector reform in more than 60 countries (FAO [1]; Kulkarni and Tyagi [10]). In IMT, it is common to set Joint System Management (JSM) that allots the management of main canals, main drainages, and main roads to the government [or private enterprise, i.e., Scheme Management Entity (SME)] and the management of lateral canals and below to the WUAs (Hatcho and Tsutsui [5]; Sato et al. [16]; Kulkarni and Tyagi [10]). However, the performance of PIM and IMT has generally been poor. It is said that WUA establishment is not easy, or established WUAs cannot function well, or Irrigation Service Fee (ISF) collection rate gradually decreases (Ishii and Sato [6]; Kakuta [7]). There are many reports of illegal gates operation and destruction and insufficient communal work for canal maintenance that lead to an unequal water distribution and then, a decline of the irrigation efficiency and the deterioration of irrigation facilities (Sato et al. [16]).

This study examines the problems of irrigation management, especially those of JSM, based on a case study in WUA AK/C1 (the official name: hereafter WUA C1) at the Kpong Irrigation Scheme (KIS) owned by the Ghana Irrigation Development Authority (GIDA) in Ghana, using the theories of common-pool resources (CPR) management proposed by Ostrom ([14]) and Freeman ([2], [3]).

The rest of this paper is organized as follows. First, the analytical framework and methodology of the study are proposed. Next, the irrigation system and the WUA are described. Then, the performance of the irrigation is evaluated. Finally, in the conclusion, the appropriateness of JSM is considered.

Evaluating the management of an irrigation system needs an appropriate analytical framework. Since the irrigation water and irrigation systems that deliver water to beneficiary farmers are considered to be CPRs, which are commonly managed by local people, the author applies Ostrom's ([14]) design principles of long-enduring CPRs and Freeman's ([2], [3]) distributional share system model to assess performance.

According to Ostrom ([14]), although there are differences among CPR settings, long-enduring and self-governing CPR institutions (e.g., irrigation system, community forest, common pasture) share eight design principles (see Table 1). (Ostrom was awarded a Nobel Prize for economics in 2009 because of this research on CPR management.). If a CPR institution does not share these eight design principles, it cannot avoid free riders who break its rules and appropriate resources unfairly without fulfilling their obligations as members of the institution. As a result, deterioration and dysfunction of the CPR institution ensues, finally leading to the depletion and destruction of the resources themselves. In the case of irrigation systems, deterioration of irrigation facilities and water shortage at the tail portion are caused, and consequently, irrigation projects end in failure. Hence, avoiding free riders is indispensable to sustainable CPR management (Ostrom [14]).

Ostrom's design principles illustrated by long-enduring common-pool resources (CPR) institutions. Source: Ostrom (1990: 90)

Meanwhile, Freeman's distributional share system model is formulated to apply to irrigation systems and WUAs. However, its conceptional model is equivalent to Ostrom's design principles. Freeman ([2], [3]) presents six essential characteristics of an effective WUA (see Table 2).

Freeman's six essential characteristics of an effective WUA. Source: Freeman (1989: 25), amended by author based on Lepper (2007: 50) and Freeman (2009) Personal conversation by e-mail

Freeman ([2]) states that a sense of fairness must be shared among WUA members to manage an irrigation system sustainably in the long run. Hence, the six characteristics in Table 2, especially a distributional share system (the third and fourth characteristics), are essential. A distributional share system has three aspects: (1) share of water, (2) share of cost, and (3) share of vote (Freeman [3]). According to Freeman, there are three conditions making WUAs successful. First, each member's share of water should be equivalent to his/her share of cost; that is, the water volumes received by each member are roughly proportionate to the share of system costs paid by each member (Freeman [3]). A WUA should have a management rule that if a member receives more benefits (e.g., water in a timely way) than do other members, he/she must pay more management costs (e.g., via an ISF or in labor, materials, etc.). If the member receives less water than others, his/her management cost should be less [Freeman ([4]) Personal conversation by e-mail]. Second, a WUA should remove head and tail distinctions in the service queue; that is, it provides the same volume of water per unit area in the command area of the irrigation system. Third, conflicts in a WUA are resolved based on each member's share of vote: If a member's share of cost is larger, his/her share of vote in the WUA will also be larger (Freeman [3]).

There are case studies of successful irrigation systems that have distributional share systems and Freeman's ([2], [3]) essential characteristics of effective WUAs (Freeman [3]; Maass and Anderson [12]; Martin and Yoder [13]; Siy [17]; Kakuta [8]). This study assesses the performance of WUA C1, using Freeman's essential characteristics of effective WUAs and Ostrom's design principles. Among Ostrom's eight design principles, the second (congruence between appropriation and provisional rules and local conditions) is considered to be equivalent to Freeman's third essential characteristic; namely, that water delivery is dependent on the fulfillment of organizational obligations (i.e., the distributional share system).

The author uses field data gathered from surveys of various stakeholders involved in WUA C1 in KIS for 3 weeks in August and September 2017. The surveys were conducted using Rapid Rural Appraisal, especially semi-structured interviews, based on anthropological research methods. The interviews asked about respondents' activities in the WUA and their farm and social settings as well as their social and family relations. The interviewees consisted of the 11 officials of WUA C1, two members of the Interim Committee of WUA in KIS, two officials of the KIS Office, and two officials of GIDA Headquarters. Field data were qualitative rather than quantitative since they were basically collected through semi-structured interviews. Hence, the performance of WUA C1 was evaluated by interpreting those qualitative field data.

The Kpong Irrigation Scheme (KIS) irrigated about 3000 hectares, 1500 hectares for rice farms, and 1200 hectares for banana farms (MASAPS-KIS [20]). KIS was a gravity irrigation system with Kpong Dam as a water source. Irrigation facilities consisted of a main canal, branch canals, and lateral canals, as well as main drainage, branch drainages, and lateral drainages along with access roads (see Figs. 1, 2).

Graph: Fig. 1Map of the Kpong Irrigation Scheme (KIS). Source : MASAPS-KIS, JICA

Graph: Fig. 2Map of WUA C1 and the Kpong Irrigation Scheme (KIS). Source : MASAPS-KIS, JICA

In KIS, canals reached to each WUA and each farm; i.e., each WUA had an inlet from the main canal to a branch canal that irrigated the whole irrigation area of the WUA. The branch canal had inlets to lateral canals that irrigated each Block (irrigation unit) under the WUA. The lateral canal had inlets to each farm. Hence, if water management was properly performed, each WUA, each Block, and each farm could manage its irrigation facilities independently within its own irrigated command, without influence by neighboring WUAs, Blocks, or farms. In developing countries, there are many irrigation systems that deliver irrigation water from a common inlet to many farms by paddy-to-paddy irrigation, and the beneficiary farmers suffer from water shortages and water conflicts as a consequence. From this viewpoint, KIS had an excellent irrigation system.

The number of beneficiary farmers in August 2017 was 2611 (MASAPS-KIS [21]); they dwelled in Akuse, Asutsuare, Oswem, and other neighboring villages of KIS. The main ethnicity of beneficiary farmers in KIS was Osudoku (76%), Ewe (19%), Akan (3%), etc. (MASAPS-KIS [19]).

KIS started an irrigation scheme in 1998. From 1998 to 2016, KIS was managed by JSM whereby the KIS Office of GIDA operated and maintained the main canal, main drainage, and roads along the main canal (hereafter called main roads), while the Osudoku Agriculture Cooperative Society (OACS), a cooperative group composed of beneficiary farmers of KIS, operated and maintained 17 branch canals and below [branch canals, lateral canals, branch drainages, lateral drainages, and access roads (hereafter called branch roads)]. Tasks of OACS were mainly the financing agricultural credits to farmers and management of irrigation facilities.

At the time of OACS's management, there were many problems, such as a low collection rate of Irrigation Service Charges (ISC), which resulted in poor maintenance and rehabilitation of irrigation facilities because of the lack of funds and led to water shortages at the tail portion of irrigation system; a lack of farmers leadership that made it difficult to organize communal labor work for canal cleaning and road maintenance; and a lack of communication among farmers that caused water conflicts. There were no regular meetings to discuss water distribution planning in KIS (MASAPS-KIS [20]).

However, a new law, L.I.2230 (WUA law), was enforced on May 2016. This law decreed that the beneficiary farmers should found WUAs and manage the branch canals and below in the irrigation schemes that GIDA had managed [Yamauchi ([22]) MASAPS-KIS, GIDA-JICA (unpublished paper)]. Hence, GIDA decided to stipulate a new management plan at KIS from December 2016. It was decided that OACS would be dissolved and that 12 new WUAs should be newly established (MASAPS-KIS [21]). Under the new management plan, 12 WUAs would operate and maintain each branch canal and below on behalf of OACS, while the main canal, main drainage, and roads were expected to be operated and maintained by a newly established private enterprise called a Scheme Management Entity (SME). The KIS Office was expected to supervise the entire OM and management of KIS (MASAPS-KIS [20]).

In October 2016, the Interim Committee of 17 branch canals was founded for the establishment of 12 WUAs. In KIS, the Ghana Commercial Agricultural Project (GCAP) under the World Bank collaborated with the Project for Enhancing Market-Based Agriculture by Smallholders and Private-Sector linkages in Kpong Irrigation Scheme (MASAPS-KIS) under the Japan International Cooperation Agency (JICA) to support the founding of WUAs (MASAPS-KIS [21]). In September 2017, the upstream WUA C1 and WUA C2 had been already founded as the Pilot WUAs, while the remaining ten WUAs were supposed to be founded by the end of 2017. However, the nature of SME was still unknown as of September 2017.

Under the new management system, the expected roles of a WUA were water management of the branch canals and below, cleaning of the branch canals and below, collection of the ISC from WUA members, holding of WUA General Assemblies, holding of WUA Management Committees, holding of Block Meetings at each Block of the WUA, conflict settlement inside the WUA, and coordination with higher bodies, such as the Interim Committee of 12 WUAs in KIS, KIS Office, and SME.

WUA C1 was designated as a Pilot WUA for the establishment of 12 WUAs in GCAP (World Bank) and MASAPS-KIS (JICA) in 2016. Since then, many activities have been conducted to facilitate the establishment of WUA C1. Since WUA C1 was located at the most upstream area of KIS, it was advantaged in terms of water allocation in KIS. The irrigated area was 107 ha, and the number of beneficiary farmers was 88 households. There were subdivisions within WUA C1, namely Blocks (Lateral) 1, 2, 3, 4, 5, and 6 (MASAPS-KIS [21]) (see Fig. 3). Under the Blocks, there were Sections A, B, C (and D at Block 5). The fact that the irrigation facilities of not only C1 but all of KIS had a decentralized organizational structure was a superior point in the design plan of the irrigation system.

Graph: Fig. 3Map of WUA C1. Source : MASAPS-KIS, JICA

The branch canal level (i.e., below the main canal level) was managed by the C1 Provisional Management Committee (PMC) in September 2017. At first, a Founders' Committee was established in February 2017. Then, at a Founders Meeting on July 10, 2017 that was attended by 67 of the 88 WUA members, there was an election of the 11 PMC members. One member came from Block 1, while two members each came from Blocks 2, 3, 4, 5, and 6 (MASAPS-KIS [21]).

Basically, each Block (Lateral) had five lateral representatives. The Block Chairman served simultaneously as a PMC member. Each lateral representative was basically elected from each section. Some of them served as section leaders of their own sections.

Above WUA C1, there was an Interim Committee consisted of the representatives of 17 branch canals.

As of September 2017, the motivation of the chairman and other PMC members was high, as WUA C1 had been just organized. Since WUA C1 had received many assistances from MASAPS-KIS, JICA, and GCAP, farmers were encouraged to manage their WUA by themselves. Moreover, communication between the KIS Office and WUA C1 had much improved compared with the time of OACS.

Based on the 13 evaluation indicators derived from the Freeman and Ostrom models, the organizational performance of WUA C1 was as follows (Table 3).

Evaluation of the organizational performance of WUA C1. Source: Author's survey

Consequently, nine indicators showed that the management of WUA C1 had improved compared to the time under OACS management, although one indicator (No. 7) became worse. Particularly, 'source of leadership' (No. 1) and 'members' support to WUA' (No. 6) were evaluated as 'Yes', since all leaders were local people, and the PMC members and WUA members were highly motivated. Although the other seven indicators showed more improvement than before, they were evaluated as 'weak', due to the existence of various problems. However, these indicators could be evaluated as 'Yes' if problems were resolved through the establishment of WUA activities and WUA management capacity was strengthened. As space is limited, hereafter, this study focuses on the three indicators (Nos. 3, 5, and 12) that showed problems.[1]

Meanwhile, WUA C1 lacked the distributional share system indicator (No. 3) that Freeman ([2], [3]) points out as the heart of any effective WUA and successful irrigation management.

In KIS, ISC was set as 260 GHS per hectare per year. Since this ISC was a fixed rate per area, it was an unfair system for farmers who experienced water shortage downstream. At the time of OACS management, the ISC collection rate in the entire KIS varied from 10.85 to 105.56%, and the average rate was quite low at 40%. Meanwhile, the ISC collection rate in Branch C1 was higher than the average rate in KIS (91.1% in 2014, 86.0% in 2015, declined to about 40% in 2016), probably reflecting the sufficient water supply at Branch C1 area.

However, there were still some members who did not pay the ISC because of the inadequate ISC collection system. At the time of OACS, most of the farmers paid the ISC in rice to the OACS Task Force which consisted of the executive members of OACS at a dry floor of OACS. Then, the Task Force paid the collected ISC to the KIS Office. However, since OACS did not issue the receipts of ISC payments, some farmers who did not trust OACS directly paid the ISC in cash to the KIS Office. There was no follow-up of ISC collection by OACS for the farmers who did not show up at the dry floor, and consequently, the free riders were just left alone. After WUA C1 was organized, farmers were newly requested to pay the ISC directly to the bank account of WUA C1. To avoid non-payment of the ISC, the PMC members planned to request farmers to show a receipt of ISC payment before land preparation, and if the farmer could not do so, PMC would take over his/her tractor until the farmer paid the ISC. However, this sanction did not seem effective because it would be possible for the farmer to simply rent another machine.

Although there were free riders who received water without fulfilling their duties of ISC payment, no effective sanction existed to avoid those free riders.

Furthermore, WUA C1 lacked a water resources control ability indicator (No. 5).

Freeman ([2]) points out that if a WUA has high water resources control ability, it can deliver sufficient irrigation water in a timely manner to all members of the WUA.

As mentioned above, KIS had been managed by JSM, in which the KIS Office managed the main canal, main drainage, and main roads, while WUAs managed branch canals, branch drainages, branch roads, and below.

WUA C1 conducted communal labor two times per year before the farming season to clean branch and lateral irrigation and drainage canals. Generally, WUA C1 could fairly maintain the branch and lateral canals through such communal labor.

To maintain branch drainage and lateral drainages, WUA C1 members sprayed chemicals and cut weeds jointly. Again, C1 could maintain the drainages fairly by communal labor.

As mentioned above, at C1, since each Block had its own inlet from the branch canal, and each farm plot had its own inlet from the lateral canal, it was possible to engage in independent and simple water management at each Block and farm level. Specifically, each farmer could take water directly from the lateral canal to his/her farm plot when he/she needed water, without being affected by neighbors, if there was no water shortage.

Meanwhile, the KIS Office could not fully operate and maintain the main canal, main drainages, and main roads because of a limited KIS Office budget.

The KIS Office could not clean the main canal in 2015 and 2016 because of a lack of fund. Consequently, sediment had accumulated at the bottom of the main canal and grass was growing inside the main canal, which caused water shortage at the downstream portion. Hence, all of the KIS beneficiary farmers of 17 branch canals groups had to arrange extraordinary communal work to dredge mud and cut grass in the main canal with human labor twice in 2016 and August 2017. While most of the farmers attended at C1 Branch, at C6 Branch, only half of the famers came, as they claimed that cleaning the main canal was the duty of the KIS Office. Because of this poor maintenance of the main canal and the hard, voluntary works of cleaning the main canal, some farmers refused to pay the ISC, and hence the ISC collection rate decreased from 86.0 to 40% at C1 Branch in 2016.

It was said that the main drainage canal had not been dredged for 8 years because the KIS Office could not arrange the heavy equipment needed because of its budget shortage. This caused the drainage canal to overflow at Block 2 and Block 4 in C1. Moreover, in June 2017 drainage water overflowed from a drainage canal and damaged a farm just before rice harvesting at Block 6 Section B. The owner of the farm refused to pay the ISC.

Further, on some portions of the main roads, vehicles could not pass because the KIS Office could not arrange heavy equipment for road maintenance. Consequently, after the rice harvesting, the farmers had to hire laborers to carry rice bags from their farms to a point of a main road where cars could enter. This caused extra labor costs for the farmers. Further, the KIS Office could not repair broken facilities, such as irrigation canals, drainages, and gates, because of insufficient staffing and budget.

Regarding the water management at the main canal level, the operation of gates from the main canal to the branch canals was handled by the staff of the KIS Office. Because of restructuring of the KIS Office, the number of water management staff was decreased from 10 in 2014 to 1 in 2017. Hence, it took 2 days to monitor the main canal and gates throughout all of KIS, and thus, the staff could not respond quickly to farmers' requests to open or close the gates or deal with overflows and other problems in the canals. Hence, complaints from farmers increased. Further, although the number of illegal operations of gate at main canal decreased after WUA C1 was founded, some farmers still opened the gates illegally or broke their keys.

At WUA C1, management of the irrigation facilities that WUA C1 took charge of seemed mostly fine. However, management of the facilities that the KIS Office took charge was inappropriate because of the limited KIS Office budget. Consequently, WUA C1 could not deliver sufficient water in a timely manner to members. The JSM arrangement in KIS reduced WUA C1's water resources control ability.

Ostrom ([14]) states that the rights of appropriators to devise their own institutions are not challenged by external governmental authorities (see Table 1).

As mentioned above, WUAs in Ghana had the right to organize which was guaranteed by the Government under Law L.I.2230 in May 2016. GIDA ordered the dissolution of cooperatives in the country and the establishment of WUAs in September 2016 [Yamauchi [22] (unpublished paper)]. Hence, WUA C1 was founded legally authorized by the WUA Law.

Compared to OACS, WUA C1 covered a smaller area so that it could concentrate on the management of its own branch canal. Further, while OACS took charge of many tasks that included agricultural credit, the tasks of WUA C1 focused on water management. Consequently, PMC members could manage the organization of their WUA more easily compared with OACS.

Since WUA C1 needed sufficient financial management capacity, it had a plan to increase the ISC from GHS 260 to GHS 553 per hectare. Also, WUA C1 needed to increase the ISC collection rate.

Meanwhile, WUA C1 needed to have the ability to introduce outside resources that a WUA ordinarily could not afford by itself from the government, private enterprises, or donors. Through the WUA establishment activities, the relationship between farmers and KIS Office had much improved. Since they had more opportunity to communicate, they gained a better understanding of each other's situation.

However, the KIS Office then underwent restructuring. The reduction of the KIS Office's budget created the following big problems for WUA C1 and the other WUAs.

As mentioned above, the lack of funds for operation and maintenance of the main canal and drainage, especially in 2015 and 2016, caused water shortages and flood and forced farmers to clean the main canal themselves. The lack of funds for maintenance of the main roads caused bad road condition, creating a situation in which farmers could not access their fields by vehicles and instead had to hire laborer to carry harvested rice. The reduction in the number of water management staff to monitor canals and gates led to many complaints from farmers.

The number of extension workers at KIS was reduced to only one person. The extension worker became so busy that he could only deal with the farmers who contacted him. Previously, the KIS Office had leased farm machineries such as tractors and combines to the farmers. But now the machines were old and could not be leased. A shortage of heavy equipment, such as tractors and combines, caused delays in such farming activities as transplanting and harvesting. Previously, the KIS Office had contracted with seed-growing farmers to produce and distribute good rice seeds. However, they stopped the system of seed growers because of the limited budget, and as a result, the farmers in KIS had difficulty getting access to good seeds. Some farmers purchased seeds from private companies, but there were problems with low seed quality that included non-germination. Some farmers claimed that their rice yield had been 7 tons per hectare before but was now decreased. Although the KIS Office had owned a dry floor and a rice storage before, they were closed now. Hence, when it rained, the harvested rice could not be dried well. Poor post-harvest quality decreased the value of the rice. At the time of OACS, there had been regular meetings with the KIS Office and farmers. However, the KIS Office has since stopped them.

Hence, WUA C1 had an organization in which PMC members could engage in management more easily compared with OACS. Meanwhile, although WUA C1 needed to introduce outside resources that a WUA cannot ordinally afford by itself, it could not access the technical assistances and financial supports that had previously been provided by the KIS Office. This caused severe problems in WUA C1 management, and the rights to organize were hampered.

Based on the evaluation of organizational performance, WUA C1 lacked the distributional share system indicator. WUA C1 could not avoid free riders who took water without performing their organizational obligation (i.e., ISC payment).

Furthermore, WUA C1 lacked the water control ability indicator. As of September 2017, since KIS had been managed by JSM, the KIS Office had to operate and maintain the main canal, main drainage, and main roads, while WUA C1 had to operate and maintain the branch canal, branch drainage, branch roads, and below. However, the KIS Office could not fully manage the main canal, main drainage, and main roads due to a lack of funds. It could not clean the main canal and drainage canals. Further, its water management staff was reduced due to GIDA's restructuring. Hence, the KIS Office could not respond to complaints and requests from farmers in a timely manner. Consequently, WUA C1's water control ability deteriorated even though it had fine irrigation facilities that enabled independent water management within the WUA.

The KIS Office's lack of funds also harmed the indicator of minimal recognition of rights to organize within WUA C1. Although WUA C1's founding was legally authorized by the WUA Law, it successfully increased its management capacity with the high morale of PMC members, and furthermore, the relationship between farmers and KIS Office had much improved, WUA C1 had difficulty accessing the outside resources and assistances that the KIS Office had previously provided.

Consequently, to have successful management of KIS, a WUA first needs to devise effective sanctions to avoid free riders in order to increase the ISC collection rate. Secondly, it is necessary to further promote IMT so that a WUA can manage its irrigation system effectively by itself. Hence, the establishment of a strong WUA is needed. If a WUA can secure sufficient financial capacity, it can hire WUA staff (e.g., accountants, bookkeepers) and pay salaries to them; purchase farm equipment (e.g., hand tractors, combines) and dry floors and lease them to WUA members; purchase farm inputs jointly at a low price; sell rice jointly at a high price; build WUA rice storages; build WUA offices and purchase office supplies; etc. Moreover, the WUA can hire water management staff (e.g., water tenders, irrigation technicians) to conduct daily operation of gates and canal monitoring to improve water management and avoid water shortages and water disputes among farmers. Further, after all WUAs are established at each branch canal area, it will be desirable to raise the Interim Committee in status to a WUA Federation in order to gain more bargaining power and to engage more in the management of the entire irrigation scheme. Thirdly, good management of the main canal, main drainage, and main roads is essential. Fourthly, a WUA needs to be able to access outside resources that it cannot afford by itself. Some services, such as leasing farm machineries (tractors, combines), seed production and distribution, and drying and storing rice, can be provided by the private sectors. On the other hand, privatizing the management of irrigation facilities (main canal, main drainage, main roads) and agricultural extension services seems to be problematic.

Under the plan devised by GCAP and MASAPS-KIS, a Scheme Management Entity (SME), a new private enterprise, will manage the main canal, main drainage, and main roads. (The KIS Office is to supervise the entire OM and management of KIS.) SME is expected to take charge of irrigation services, support for input supply, extension services, and support for processing and marketing, which the KIS Office provided previously (MASAPS-KIS [20]). However, as of September 2017, although there was an expectation that Golden Exotics Ltd. (a banana production enterprise with 1200 hectares of banana farm in KIS) would take on this SME role, it was still uncertain whether an adequate SME can be found and whether the SME can really manage the main facilities (main canal, main drainage, main roads) of KIS on behalf of the KIS Office.

In IMT, it is common to set JSM. To make JSM successful, it is necessary to have strong WUAs and simultaneously enough support from a government (or SME) that can provide the necessary resources in a timely manner, including water resources from the main canal, to the WUAs. Consequently, there must be a good and close relationship and communication between the WUAs and the government (or SME); moreover, the government (or SME) should have enough manpower, funds, and technical skill to support the WUAs. Otherwise, from the perspective of evaluation of the organizational performance of WUAs, it is possible that JSM can harm the water resources control ability of WUAs.

In the case of KIS, there was a vacuum in the management of the main canal, main drainage, and main roads that the government side was responsible for. In KIS, the problem of inability to control water was not caused by the beneficiary farmers (WUA) but by the government. However, the possibility of SME-led management still seems ambiguous. In Ghana, there were two cases in which SMEs took charge of the management of main irrigation facilities on behalf of GIDA (and WUAs took charge of branch canals and below). However, both failed in the financial management of the irrigation scheme because they could not collect enough ISC from beneficiary farmers. Since a SME is a private enterprise, it will withdraw from managing an irrigation scheme if it cannot secure enough profit. The withdrawal of a SME from irrigation management will create a calamity for the beneficiary farmers.

One possible option would be for the above-mentioned WUA Federation to take charge of the operation and maintenance of the main canal, main drainage, and main roads of KIS in the future, if the Federation can function effectively. While the SME might impede the formation of the nested enterprises that Ostrom ([14]) points out as a necessary design principle, the WUA Federation is more suitable for securing consistency in the management of KIS in its entirety, from the water source (dam) to the end of canals (farm), by forming the nested enterprises of the WUA Federation, WUA, Block, section, and each farm in KIS.

The author wishes to thank the Japan International Cooperation Agency (JICA), the Ghana Irrigation Development Authority (GIDA), and all concerned in the Kpong Irrigation Scheme (KIS) for their cooperation with the author's survey and research. Also, the author wishes to thank Dr. David Freeman, Professor Emeritus, Colorado State University, for his valuable comments and suggestions.