Management Models and the Sustainability of Rural Water Supply Systems: An Analytical Investigation in Ha Nam Province, Vietnam

Abstract

Clean water and sanitation is a global issue that has recently received special attention from governments and international organizations, as they essential to enhance people’s health and foster socio-economic development in different countries. Therefore, one of the United Nations’ Millennium Goals is to ensure access to safe drinking water for all as its target by 2030. However, ensuring water supply in rural regions, especially in developing countries, is a challenge. To meet the target, investing in infrastructure and maintaining the sustainability of water systems is crucial. This study takes into account the role of the management model of rural water supply systems and its relationship with the sustainability of the water systems. A weighing procedure for a sustainability assessment of rural water systems for the current study was proposed, then an ordered logit regression model was applied to examine the relationship between types of water system management models and the sustainability of the water systems in the context of Vietnam. The results from this study showed that type of management model had an influence on the sustainability of rural water supply systems, and the private management model was considered the one that significantly contributed to the sustainability of the rural water systems in the study area. The study provides some lessons learned for researchers in the field of rural water supply to select appropriate approaches to assess the sustainability of water systems and for policymakers to modify current policies toward more sustainable development of water infrastructure in the future, especially in developing countries.

1. Introduction

Water is essential for all creatures on earth, and it has been suggested to be a human right [1]. However, millions of people still have not been able to access clean water all around the world. According to the United Nations International Children’s Emergency Fund (UNICEF) and The World Health Organization (WHO) data, by 2017, there were 875 million people who could not access basic drinking water services, most of whom were living in developing and under-developed countries [2]. The COVID-19 pandemic has also highlighted the need for everyone to access clean water and good hand hygiene facilities [3]. Because of the importance of water, the United Nations has set water-related targets in its Sustainable Development Goals (SDGs) to be achieved in the 2030 Agenda [4]. These goals include targets to reduce water, sanitation, and hygiene inequalities progressively since water supply plays an important role in ensuring people’s sanitation and hygiene [5]. Specifically, Goal 6 aims to ensure availability and sustainable management of water and sanitation for all. Therefore, the supply of water for domestic uses to rural areas, harmonization of interests between the parties in supply and use, and ensuring fairness and sustainable development are key to achieving by policymakers [6]. In practice, water infrastructure is multifaceted and plays an essential role in achieving the goals [7]. Nevertheless, obtaining the sustainability of rural water supply is not an easy task due to the complexity of the issue [8]. The term ‘sustainability’ in this case means that the water supply systems can be maintained for long-term operation and supply water to its customers on demand.

Because of the water sector’s distinctive characteristics, rural water supply systems have been built and managed by various models. The most common types of management models are structured under community-based, public service-based, and private-based forms [9]. In practice, each type of management model has its unique features depending on the specific context of countries; therefore, it affects the water supply. The key difference of each type originates from the objective of water supply activities. For example, the objective of the private-based type is driven by profit; meanwhile, that of the community-based form is for the benefit of the whole community. The management model can affect various aspects of the water supply, such as social, economic, technical, or managerial perspectives. The sustainability of rural water supply and its services has been widely discussed in the literature, including the consideration of its impact factors (for example, see [8,10,11,12,13,14,15,16,17,18,19,20]). However, the use of quantitative analysis to evaluate the relationship of different types of management models with the sustainability of rural water supply systems is absent.

This study takes into account the effect of management models on the sustainability of rural water supply systems. The aim is to examine whether the type of management model of water systems plays a role in contributing to the sustainability of the systems by adopting a proposed procedure to test it in practice. To achieve this, first, it reviews the literature on the use of different models in managing rural water supply systems as well as identifying the factors affecting the sustainability of water systems. Then, an overview of management models in the rural water supply sector of Vietnam was given, and a procedure to evaluate the sustainability of water supply systems in the study cases was undertaken. An econometric model was established to examine the relationship between different types of management models and the sustainability of the water systems. The analysis model was based on the data collected from the investigation of rural water systems in Ha Nam province in the Red River delta of Vietnam. The study results in helping to clarify the role of the management model in its relationship with water system sustainability and suggest policy changes to promote more sustainable rural water supply models in the study context. The paper ends with some conclusions.

2. Management Models and Sustainability Assessment of Rural Water Supply

2.1. Management Models of Rural Water Supply

Much research has been undertaken focusing on models of rural water supply. The literature in the field of rural water supply shows a variety of models for the management of rural water systems, and it depends on the situation as the development of the economy as well as the specific context. Moreover, different models may exist at a certain point in time within the same country. A number of models that exist in most countries are under public service-based, community-based, private-based forms, and individuals [21,22,23,24].

With the public-based model, some arrangements of operation can be shown, such as:

• Direct public management, in which water supplies are operated directly by public authorities.
• Delegated public management, in which supplies are operated by not-for-profit entities which are appointed or created by the public authorities to run the supply.
• Delegated private management (known as public–private partnerships), in which supplies are operated by private (for-profit) companies, subcontracted by public authorities, often through a lease or concession contracts. With the private-based model, management is not common in rural areas, and it is normally done by direct private management.

There are numerous advantages of this model, such as the quality of water services could be better ensured, and the subsidy mechanism can be easily specified. In addition, operation workers of this model are often trained with technical skills; therefore, water systems are properly operated and maintained. However, this model requires routine support from the government, and in case of state budget limitations, problems may arise.

In the community-based model, the water supply is managed by community members, and it also involves the participation of civil society organizations. This model has a number of advantages and disadvantages. For example, this model creates proximity to water users, and local capacity to manage conflicts and users have a better mechanism to express their demands and address their concerns. However, one major disadvantage of this model is the lack of capital and, therefore no guarantee in case of mismanagement and limited technical skills to manage water systems properly [25]. Nevertheless, this model will work best where the members of the community owning and operating the supply are linked through shared values and principles, such as voluntary and open membership; democratic decision-making; economic participation of all members; training and sharing of information among members; and concern for common water-related problems.

With the private-based model, the water supply is directly managed by private investors. The private entity could be an enterprise or an individual who invests in the water systems or is assigned the right to manage the water infrastructure. This model has major advantages, such as the financial and management autonomy of the supplier and a better understanding of the levels of service that consumers are expecting. However, this model may not support the community as the provider often does not have incentives to expand the services to all user segments in case it does not bring benefit to the investor.

Many advantages and disadvantages of different management models of rural water supply services have been discussed in more detail in several documents; for example, see Münger et al. (2009) [25].

With the household-centered model, water supply is provided by itself from its investment. In this case, water systems are usually small, with uncomplicated techniques required.

Management models of rural water supply are summarized in

Table 1. Various models of rural water supply.

Model	Scale	Investment Source	Objective	Country of Application
Public- or government-based model	Large and medium	Government budget	Non-profit; social objectives	France, Austria, Germany, Finland, Italia
Private or market-based model	Large and medium	Private enterprises/Individuals	Profit-driven	UK, USA
Community-based model	Medium and small	Members of the community/sponsors from outside	Community benefits	Finland, Austria
Household-centered model	Small	Households/individuals	Household benefits	Africa, Asia, Latin America

Source: Hendry and Akoumianaki [24].

.

2.2. Sustainability Assessment of Rural Water Supply Systems

The literature in the field of water supply has shown numerous studies on the sustainability of water supply systems. The fundamental elements that define the sustainability of water supply systems are identified, such as service availability, user satisfaction levels, water consumption, the level of payment of households for water, stability in O&M of the system, and organizational sustainability (e.g., the system’s organizational structure, leadership abilities, and human resource standards) [10,11,12,13,14,26]. Several studies emphasize the requirements that must be met in order to guarantee the water supply and use system’s long-term viability, including the Sustainability of input water sources; Sustainability of construction capability; Sustainability of management apparatus; Sustainability of technological advancement; and Sustainability of regular water availability [15,16,17]. Peter and Nkambule [15] generalized those criteria to form technical, social, environmental, financial, and organizational sustainability dimensions. To emphasize the importance of sustainability, Maryati et al. [27] stated that the sustainability issues of water supply systems must be identified to achieve the target of universal access to drinking water. Therefore, how to assess the sustainability of water systems has received special attention from scholars.

However, the sustainability of the water supply, including financial security for system operations, is largely dependent on the sustainability of the water supply system. If the water supply operator does not have sufficient funds to provide the service, accumulate capital for reinvestment, or upgrade the system in order to maintain or improve the quality of the water service, it will increase the risk to the survival of the water supplier. Lack of funds for maintenance and repair of machinery and equipment and reinvestment will result in inefficient operation and low service quality that will negatively affect water users. Poor performance of water systems can result in a vicious cycle of the water supply [28,29]. In a vicious circle, the water supply service is not guaranteed in terms of frequency and quality; people do not trust to use it regularly, then this leads to low revenue, low profit, and a lack of capital to invest or purchase necessary equipment and services; therefore, failure to implement regularly repair and maintenance of machinery and equipment, and then the water systems cannot operate efficiently and sustainably.

Studies on the assessment of the sustainability of water supply systems suggest determining criteria with certain factors, and these criteria can be generalized to form six dimensions, including (1) Institutional; (2) Managerial; (3) Social; (4) Financial; (5) Technical; and (6) Environment [15,30]. The dimensions are interacted with each other and can be shown in Figure 1.

A number of studies have also tried to identify indicators for each of the sustainability dimensions, such as participatory planning, appropriate technology choice and social marketing influencing effective community demand, local financing and cost recovery influenced by local borrowing and saving schemes, financial planning and community cross-subsidies [18]. More recently, some studies suggest determining the sustainability of water supply systems based on the criteria set using index scores [11,12,19,20,31]. This approach brings opportunities for quantifying the assessment and making it become more quantitative and comparative.

2.3. Rural Water Supply and Management Models in Vietnam

From the 1960s of the 20th century to date, there is significant progress in water supply and sanitation in rural areas of Vietnam. In order to achieve the set objectives of each period, the government has issued various policies as well as applied different approaches for implementation because there had been certain socio-economic development conditions and levels of public awareness during each period of time. In the 1960s and 1970s (a period with difficult conditions as the country was facing the war), a key solution to the issue of rural water supply was to allow people to either construct their own water sources or use existing natural water sources to improve the quality of the available water, such as drilling wells and constructing rainwater collection containers. This approach led to the construction of millions of works by people to meet their own home needs.

In the early 1980s, UNICEF in Vietnam started a rural water supply program that began with the dissemination of construction methods and partially funded the construction of small water supply systems for rural areas, such as hand-drilled wells in six provinces in the Mekong Delta [32]. The deployment was then spread to Thanh Hoa province of North Vietnam and the Central provinces by the early 1990s.

Since 1990, the nation’s socio-economic growth has significantly progressed, and the diversification of rural water supply systems has been promoted as well. Village and commune-level concentrated small-scale water supply projects had been undertaken. In this period, centralized water supply systems, distributed by pipes and water meters, were considered a long-term and sustainable solution. Many government policies were issued, contributing to promoting rural water supply systems. Some essential policies could be mentioned, such as Decision 237/1998/QD-TTg of the Prime Minister in 1998 regarding the National Target Program on water supply and environmental sanitation in rural areas. In 2000, the Government issued Decision No. 104/2000QD-TTg approving the National Strategy for water supply and environmental sanitation in rural areas until 2020. More recently, the Prime Minister in Decision No. 1978/QD-TTg, dated 24 November 2021, on the “National strategy for rural water supply and environmental sanitation till 2030, with a vision to 2045”. The overall objective of the Strategy is to guarantee that rural residents have the right to quick access to safe water and affordable water supply services. The specific target of the Strategy is for 65% of rural dwellers to have access to clean, standard-quality water by 2030. The target of 100% of people in rural areas to access reliable water supply and sanitation by 2045.

Over the last three decades, numerous rural centralized water supply systems have been built across Vietnam. The numbers and scales of rural water supply systems in seven provinces of Vietnam are given in

Table 2. Summary of the number of rural water supply systems and their scale in seven provinces of Vietnam.

Water Supply System Scale (m3/Day-Night)	Provinces
	Ha Nam		Bac Ninh		Vinh Phuc		Hung Yen		Quang Ninh		Ha Noi		Thai Binh		Long An
	No of Sys.	Perc. (%)	No of Sys.	Perc. (%)	No of Sys.	Perc. (%)	No of Sys.	Perc. (%)	No of Sys.	Perc. (%)	No of Sys.	Perc. (%)	No of Sys.	Perc. (%)	No of Sys.	Perc. (%)
≤100	4	6	0	0	14	28	0	0	156	83	10	8	1	1.2	53	3.8
>100–200	5	7	1	2	2	4	0	0	3	2	15	13	29	35.4	606	42.9
>200–500	31	44	1	2	14	28	0	0	12	6	15	13	16	19.5	681	48.3
>500–1000	10	14	10	23	3	6	20	43	7	4	54	45	17	20.7	44	3.1
>1000–1500	1	1	6	14	4	8	5	11	5	3	12	10	7	8.5	19	1.4
>1500–2000	4	6	9	20	2	4	2	4	1	1	6	5	4	4.9	7	0.5
>2000–3000	9	13	4	9	7	14	7	15	2	1	3	3	3	3.7	0	0
>3000	6	9	13	30	4	8	12	26	1	1	4	3	5	6.1	0	0
Total	70	100	44	100	50	100	46	100	187	100	119	100	82	100	1410	100

Source: Directorate of Water Resources (2021) [33].

.

Table 2 shows a variety of scales of the water supply systems in different provinces, and this indicates the fact that the scale of the water system depends on the concentration of residents in different areas. For example, the provinces of Bac Ninh and Hung Yen are two with a high-density of the population; therefore, the scale of water systems in these areas is relatively greater than those in other provinces.

Under the current situation of the water sector, five types of management models are common in Vietnam: (1) The village-level model (community model); (2) The joint stock cooperative model; (3) The Commune People’s Committee model; (4) The state-based model that is managed by the Center of Rural Water Supply and Environmental Sanitation (CERWASS); and (5) The enterprise or private model [9]. The model of provincial CERWASS differs significantly in terms of the system size and the function of the organization. Regarding this model, water supply systems are often large or medium size and the function of the organization is operated based on the state’s management mechanism.

Specific types of rural water supply models in Vietnam can be briefly described as follows:

2.3.1. Community Management Model

Small-scale water supply systems for a few dozen to several hundred households are typical in mountainous areas. The infrastructure under this model is often built from the public budget or with support from foreign non-government organizations (NGOs) to improve the quality of rural community living. These water systems frequently do not include water meters, and the quality of water services is often low. Households are typically the only ones that pay service fees, and they often pay a very small amount to the person who is in charge of controlling or operating the system. This model could work more effectively if there is close attention and coordination from the authorities and residents of the village. However, this is a low-level organizational model, and the participation of people is considered a shared responsibility in the community. When encountering major problems beyond their ability, the infrastructure is easy to become degraded. This situation is often originated from the inefficient operation and maintenance of the system after a period of time.

2.3.2. Commune People’s Committee Model

This model is relatively popular in the northern region of Vietnam. Under this model, Commune People’s Committee (CPC) is the investor and sets up a water supply project management board to carry out the construction works. After completing the project, the CPC establishes a water service team to conduct O&M activities. The CPC establishes the water service team’s operating rules, which typically specify the rate of revenue and expenditure on the total amount of water supplied at the headwork point and the valves at village branches, setting the selling price of water service to its consumers. Drafting and setting regulations are usually through public discussions between the CPC and water users with the participation of representatives of local socio-political organizations. Because the CPC typically decides to establish the service team, this model typically only achieves moderate and low levels of socialization. The formulation of the service team has not yet been done following certain rules, but it depends on the local circumstances in each area. This model is also operated without the concrete of service contracts, business registration, and sound legal status. This type of model’s operation is non-profit and generates enough income to pay the costs (usually without a return on investment as well).

2.3.3. Joint Stock Cooperative Model

This model is relatively common in Vietnam, and it operates in accordance with the Cooperative Law (currently the Law on Cooperatives 2013). Under this model, there are two different types of cooperatives in providing water services: (1) Cooperatives that are established to provide many services, including containing water services; (2) Cooperatives that are founded to provide water services only. With the first type, the level of socialization of water supply activities differs from region to region due to the flexibility of legislation in each province across the nation. In the second type, the cooperative is running its business based on providing water services to its customers.

In Vietnam, a cooperative is newly established in accordance with the Law on Cooperatives, or in some cases, cooperatives that provide water services are set up motivated by local government due to failures of the community model in managing and controlling water supply activities. The model has some features that can mobilize resources from the locals for rural water supply activities, and the government does not need to implement management functions in the water system. In general, this model can be seen as a good model to create a market for rural water services in different regions across the country.

2.3.4. Center of Rural Water Supply and Environmental Sanitation Models

The state-based model of the Center of Rural Water Supply and Environmental Sanitation (CWSS) is under the provincial People’s Committee with its functions for water supply provision via stations or boards. This model type is usually formulated for relatively large, centralized water supply systems in rural areas. In practice, activities of this type are operated under a non-profit mechanism. This model aims to promote the efficiency of exploitation management and maintain the sustainability of the water supply system. The success of this model can be attributed to the following reasons: first, the project has a relatively large scale, inter-communal boundary; secondly, water system management requires operating managers with a high level of technical qualifications, so the operation and maintenance of systems are better undertaken. However, the operation mechanism of this model is basically dependent on the interference of the state and not based on the principles of markets. The socialization of water supply activities in this model is limited and financial assistance for investment and reinvestment from the government is common.

2.3.5. Enterprise or Private Models

A private enterprise is an organization or unit that self-funds and contributes capital to invest in the construction of rural water supply systems and directly manages it. In the case of sharing funds with the government, the participation of the private sector contributes a certain amount of investment capital (the rest is supported by the state), and the private enterprise will directly implement the activities from the construction stage to the management and operation phase. This type of private enterprise is often under the name of a private limited company, joint stock company, limited company, or private company.

Another form of enterprise that operates in this type of model is state-owned companies. In this form, water infrastructure is the asset of the state, and the authorized agency appoints the assignment of the water structure for the management and operation of a state-owned company. This model is with potential to establish a rural water market as well as to promote a sustainable rural water sector. However, there are still critical distinctions between state-owned and private enterprises. With regard to state-owned enterprises, the main responsibility for emerging assets is the government, but not the enterprise itself. The enterprise is only assigned the responsibility of the management of the assets. In addition, people working for state-owned enterprises are eligible to receive benefits like other state employees, such as salaries, insurance, and pensions. Meanwhile, employees in private enterprises often have fewer rights to benefits and other incentives compared to those in state-owned enterprises.

2.3.6. Summary of Management Models

Although there are a diversity of rural water supply models in Vietnam, the application of models differs in regions across the country. Each province may consist of certain types depending on the diversity of scale as well as the functions of water systems. According to the data from the Directorate of Water Resources [33], by 2021, the country have built 16,573 rural water supply systems, providing water services for about 88.5% of the population. The lifespan of the water supply structures can be classified as: above 21 years of age, accounting for 6%; between 16–21 years, accounting for 27%; between 11–15 years, accounting for 40%; and less than 10 years, accounting for 27%. The type of gravity water supply structures occupies about 62.97%, and the type of pumping systems accounts for 35.91%.

Among the types of management models, the community management model accounts for a major part, about 50%. Next is the Commune People’s Committee model, which accounts for 28.9% (mainly in the Northern Mountains and Central Highlands). The proportion of water systems managed by state-owned enterprises, private enterprises, individuals, and the CWSS accounts for approximately 10%. However, in the delta regions, the majority of water systems are managed by state-owned enterprises, the private sector, and the CWSS. The model of enterprises and private sector accounts for a major part of the Red River delta (accounting for 66.8%), whereas, in the Southeast and the Mekong River delta, the model of the CWSS accounts for a majority (approximately 40%) [33]. The proportion of management models for rural water supply systems in Vietnam by 2021 is shown in Figure 2 below.

The characteristics of rural water supply systems play an important role in influencing the sustainability of water services. In the case of Vietnam, sustainability could be affected by several factors such as water depletion, degradation, and pollution; lack of technically guaranteed capacity; high cost of producing and distributing water; lack of capital to meet the needs of upgrading and repair of structures; and limited management capacity, which results in damaged and degraded structures. In this context, the management model is also considered to play an important role in affecting the sustainability of rural water systems. This study investigates the rural water supply systems in Ha Nam provinces and applies a technique of regression analysis to solve the given research problem.

3. Study Area

The study area is Ha Nam province, which is located in the Red River delta of Vietnam (see Figure 3). Ha Nam borders Hanoi to the south as well as the other rural provinces of Hung Yen, Thai Binh, Nam Dinh, Ninh Binh, and Hoa Binh. It has a total natural area of 862 km². The population of the province is 944,200 thousand people, of which 71.73% of the population lives in rural areas (accounting for 672,126 people) [34]. The province has a significant population concentration with high demand for water services. By June 2021, 96.5% of rural households could be able to access clean and hygienic water, of which 85% are able to access clean water from centralized water supply systems. The target of the province is to develop more centralized water supply systems to provide clean water services to 100% of the population by 2030.

Regarding the natural and socio-economic conditions of the study area, Ha Nam has abundant surface water sources; however, these sources are vulnerable to being polluted by different factors such as domestic, industrial, and livestock waste. Most of the province’s water supply systems obtain their raw water from the Chau and Sat rivers. Since these rivers have a high risk of pollution, most water supply systems have built sediment and storage reservoirs to ensure the quality of raw water as a good input for treatment. A number of processes for biological, physical, and chemical treatment have been strictly followed to make sure to meet the standard of clean water before it is supplied to the user.

The data collection and investigation were undertaken in 2019. Secondary data on water supply systems were collected from the Ha Nam Department of Agriculture and Rural Development, Provincial Department of Planning and Investment, Provincial Department of Finance, District Division of Agriculture and Rural Development, and Communal People’s Committees managing units of water systems. Primary data, mostly about the evaluation regarding the sustainability of water supply systems and policy recommendations, were collected from relevant state management units in the area and direct managers/operators of the water systems. A total number of 52 systems were investigated with relevant data that were used to be examined for the purpose of the analysis. As given in Table 2, the total water supply systems in Ha Nam province were 70. Thus, the number of systems investigated in this study of 52 over 70, representing 74% of the total systems of the region.

4. Methodology

In order to answer the question of whether there is an effect of the management model on the sustainability of rural water supply systems in the current cases, a process has been adopted; first, a method was applied to assess the sustainability of water systems, then, an Ordered Logit regression model was used to determine whether there is a relationship between management model and the sustainability of water systems based on the data collected from the case studies.

First, the sustainability of water supply systems can be assessed based on the criteria that have been proposed by a number of studies (for example, see [11,12,15,19,20,31]). In this study, six dimensions were examined based on specific characteristics of water systems as well as the socio-economic context of Vietnam. These are described as follows:

• Institutional sustainability: The management model is fitted within the state’s regulations, and its business activities are accepted by the community.
• Environmental sustainability: The water source is not over-exploited and is replenished naturally over time. The water is not polluted, and its quality meets input water standards (given by regulations of the government).
• Financial sustainability: All operating and management costs are covered by the revenue from the water supply.
• Social sustainability: The processes of planning, design, and operation processes are with the participation of the community.
• Technical sustainability: The water system is well operated and maintained, and it provides water meeting given standards.
• Managerial sustainability: The management apparatus is competent and well-structured for operation and management activities.

Table 3. Specific dimensions for assessing the sustainability of water supply systems.

No	Dimension	Requirement	Key Indicators
1	Institutional sustainability	Adoption of the state’s regulations on service providers and production and business activities receive the state’s incentives. Policies and regimes (if any) and be accepted by the community.	Organize management appropriately with the current legal framework. To be eligible for preferred State policies for rural water supply (policy on capacity training, financial support) (if any). The community accepts business activities and water supply services.
2	Environmental sustainability	The water source for the system is reliable in both quantity and quality, avoiding overuse and conflicts with other water users.	Surface water: be able to ensure a coming water source with enough quantity, especially in the dry season. Groundwater: enough water reserves for extraction. Water quality: assessed according to the criteria of surface water or groundwater as input for domestic supply.
3	Financial sustainability	The revenue from water provision is enough to cover the costs of water supply and delivery.	Total revenue exceeds the cost of water provision. Ready level of funding sources. Level of community acceptance to pay a water price. Financial accounting is clear and transparent.
4	Social sustainability	Community participation in project planning, design, supervision, and operation.	The community’s high consensus in project formulation, design, and funding contribution. Understanding of water supply and hygiene sanitation.
5	Technical sustainability	The system is well managed and operated and can provide water with given standard levels.	Any system failure incidents should shortly be fixed. Routine maintenance of the system is well undertaken. Good conditions of operation and maintenance. The quality of output water meets the given criteria. The rate of water loss is modest.
6	Managerial sustainability	The operating system is proper, has enough capacity, and is authorized by the legal and technological systems.	Management apparatus is properly trained and technology transfer. Appropriate structure of management model. Having belief of the community.

Source: Modified and updated from various studies by the authors.

describes the specific indicators with respect to each of the dimensions for assessing the sustainability of water supply systems in the study area.

In practice, various methods can be used to assess the sustainability of rural water supply systems, such as modeling, muti-criteria analysis, and weighted scoring method. In this study, a weighted scoring method was adopted. This method aims to evaluate the sustainability of water systems by scoring each dimension and then summing up all the results for an overall score. Since each dimension may have a different role in influencing the sustainability of water systems, a weighted score structure was applied in this study. Before summing up the dimension scores, the weight indicating the relative importance of each dimension must be multiplied by the score.

The total score indicating the sustainability of water systems is determined by the formula:

$$E={\displaystyle \sum }_{i=1}^n{V}_i{W}_i$$

(1)

where:

• E: the total score indicating the sustainability of the water system;
• V_i: the score for assessing the sustainability of ith dimension;
• W_i: the weight of the ith dimension;
• n: number of the dimension for assessing the sustainability of the water system.

4.1. Determining the Weight

All six dimensions directly affect the long-term sustainability of water systems. In practice, the importance of each dimension can be determined by consulting experts for assessment. The importance of each dimension may differ in different contexts of institutions as well as the socio-economic development of each country. For example, the study results from Boukhari et al. [19] showed that the two most important dimensions accounting for the weight of 64.4% involve the economic dimension (38.3%) and the technical dimension (26.1%). Different approaches to assess sustainability may reflect the results from different angles. In this study, a weighting scale based on a factor score was applied. The structure of weighting scores was withdrawn from consultation with 12 experts, including researchers, policymakers, and practitioners in the field of rural water supply. The consulted results with the weighted score given to the importance of each dimension are shown as follows:

-

Institutional sustainability: a factor of 2;

-

Environmental sustainability: a factor of 2;

-

Financial sustainability: a factor of 2;

-

Social sustainability: a factor of 2;

-

Technical sustainability: a factor of 1;

-

Managerial sustainability: a factor of 1.

4.2. Determine the Score According to Sustainability Level

To determine the score for each level of sustainability, a question of agree or disagree about the scores given to four levels of sustainability to the same 12 experts above. This consultation received a high consensus from the experts regarding the points in response to a certain level of sustainability. A four-item Likert scale was used to assess the level of sustainability of the water systems, consisting of Highly sustainable; Moderately sustainable; Slightly sustainable; and Unsustainable. The total score to determine the overall level of sustainability is classified as follows:

• Level 4: Highly sustainable: 4 points;
• Level 3: Moderately sustainable: 3 points;
• Level 2: Slightly sustainable: 2 points;
• Level 1: Unsustainable: 1 point.

The score structure for the sustainability of water systems is shown in

Table 4. The total score for each dimension of the sustainability with weights.

No.	Dimension/Sustainability	Highly Sustainable	Moderately Sustainable	Slightly Sustainable	Unsustainable
1	Institutional sustainability	8	6	4	2
2	Environmental sustainability	8	6	4	2
3	Financial sustainability	8	6	4	2
4	Social sustainability	8	6	4	2
5	Technical sustainability	4	3	2	1
6	Managerial sustainability	4	3	2	1
Total		40	30	20	10

.

From the aggregate scores from Table 4, the assessment of the sustainability of each water system was undertaken based on the criteria given with respect to the result calculated from Equation (1) above as follows:

• Highly sustainable: E is between 36 and 40 points, of which there are at least three dimensions with Highly sustainable and one another with Moderately sustainable.
• Moderately sustainable: E is between 30 and 35 points, of which all dimensions must be Moderately sustainable, or four dimensions are Highly sustainable, and the two other dimensions are Slightly sustainable. None of the dimensions are Unsustainable.
• Slightly sustainable: E is between 19 and 29 points, of which there are at least three dimensions of Slightly sustainable.
• Unsustainable: E is equal to or less than 18 points.

In order to determine specific scores for each dimension of the sustainability and to assess the sustainability of each water system in the study area, a team of researchers was employed to take the tasks. At each water system, key persons from the management board were asked to make a group to discuss and give a score for their own system. The scores were recorded and summarized for analysis.

After the sustainability of water systems was determined, an Ordered Logit regression model was used to evaluate the effect of the management model as well as other examined variables on the sustainability of the rural water systems for the study area. In this case, the model is presented as a latent variable, and the method of estimation is the method of maximum likelihood.

The regression model consists of a dependent variable Y, which is an ordered variable receiving the value of j = 1, 2, 3, … J and X is a vector of the independent variables [33]. The underlying estimation principle is simple to estimate:

$$\mathrm{Pr}\left(Y_i\le j\right)=\mathrm{Pr}({\beta }_1{X}_{1i}+{\beta }_2{X}_{2i}+\cdots +{\beta }_k{X}_{ki}+{\epsilon }_i\le {\propto }_j$$

(2)

where:

• ${\propto }_j$: is a constant (or cutpoint) being satisfied with the condition ${\propto }_1\le {\propto }_2\le \dots \le {\propto }_{J-1}$
• X_i: is the vector of the independent variables with respect to observation i; i = 1, 2, … n, is the ith observation, where n is the sample size; k is number of independent variables to examine.
• $\beta$: is the vector of the parameters of the independent variables, and it describes the impact of the independent variable on the odds ratio of category j or category less than j.
• $\epsilon$: is a random error.

Equation (2) gives the cumulative probability that Y_i falls in category j and below (i.e., in category 1, 2, …, or j). Taking the log both sides of the equation and then simplify it, a new equation can be obtained as follows:

$$\mathrm{logit}\mathrm{Pr}\left(Y\le j\right)=\ln \left[\frac{\mathrm{Pr}\left(Y\le j\right)}{1-\mathrm{Pr}\left(Y\le j\right)}\right]={\propto }_j-{\sum }_{n=1}^K{\beta }_n{X}_{in}\hspace{1em}j=1,2,3,\dots ,(J-1)$$

(3)

According to statistic theory, the coefficient in Equation (3) will not be consistent if an OLS regression is estimated; therefore, the LM-maximum likelihood should be opted [33,35].

In this study, variables in the regression model are defined and measured as follows:

Dependent variable: The sustainability level of the water systems (Sustainability) that had been determined based on the data collected from investigation results from each water system. This variable involves four values with respect to the level of sustainability: Level 1: Unsustainable; Level 2: Slightly sustainable; Level 3: Moderately sustainable; and Level 4: Highly sustainable.

Independent variables: The key variables examined in the model were the types of management models of rural water supply systems. Other variables relating to the water systems were also considered, including the number of years that the system has been operated, the design capacity of the system, the investment unit rate, and the operating performance of the water system.

The management model variable (Model_type) was coded under a categorical form. In the study area, there are three types of management models, and they were arranged and assigned with a number value as the following categories: (1) Commune People’s Committee, (2) Joint stock cooperative, and (3) Private enterprise. Other variables consist of real numbers and are considered continuous. The age of the water systems (Lifespan) was determined by the number of years that the water system has been operated up to the present. The system’s design capacity (Design_cap) was the total capacity of the water system (m³/day-night) as recorded in its profile data. The operating efficiency of the system (Performance) was calculated in percentage (%) as the actual capacity divided by the design capacity. The investment unit rate of the water system (Invest_rate) was calculated by the total investment cost divided by the capacity of the system (million VND/m³). Since these values were real numbers incurred at a certain time in the past, the rates were converted to a comparative point of time using inflation rates by given years to remove the effect of inflation on the money values. The variables to be examined in the model are described in

Table 5. Key variables to be examined in the model.

Variable	Measurement	References
Dependent variable
Sustainability	Four levels of sustainability with respect to the number values of 1 (Unsustainable), of 2 (Slightly sustainable), of 3 (Moderately sustainable), and of 4 (Highly sustainable).	[15,36,37,38]
Independent variables
Model_type	Three types of management models with respect to the number values of 1 (Commune People’s Committee), of 2 (Joint-stock cooperative), and of 3 (Private enterprise).	[36,39]
Design_cap	The total capacity of the water system (m3/day-night)	[11,15,38]
Performance	Is equal to the actual capacity divided by the design capacity (%)	[11,15,38]
Lifespan	Determined by the number of years that the system has been operated to the present (years)	[39,40]
Invest_rate	Calculated by the total investment cost divided by the capacity of the system (million VND/m3)	[41]

Source: Summarized by the authors.

.

The sample size of 52 is acceptable since it meets the minimum size condition of 50, as suggested by Hair et al. [42].

The aim of using the model is to seek answers to the research question that questions if there is a relationship between the independent variables and the dependent variable.

5. Results and Discussion

Based on the procedure proposed for this study, the first 52 rural water supply systems in Ha Nam province were investigated and assessed for the sustainability of the systems. The scores were given to the six dimensions of sustainability of each system, which were summarized for analysis. The details of the specific and overall scores of all water systems with respect to each dimension are given in Appendix A. The scores of the six sustainability dimensions of water supply systems in the study area are summarized in

Table 6. Summary of scores of sustainability dimensions.

Dimension	Mean	Stand. Dev.	Min	Max	Mean/Expected Value (%)
Institutional sustainability	5.962	1.793	2	8	74.53
Environmental sustainability	5.077	1.398	2	6	63.46
Financial sustainability	5.769	2.193	2	8	72.11
Social sustainability	6.231	1.926	2	8	77.89
Technical sustainability	2.923	1.026	1	4	73.08
Managerial sustainability	3.019	1.057	1	4	75.48
Total score	28.981	8.576	10	38	72.45

:

Table 6 indicates a wide range of evaluation results with respect to each sustainability dimension from the study area. Most dimensions are with a score between the lowest value of 1 and the highest value of 4, except the Environmentally sustainable (the highest value of 3). The first four dimensions are with a factor of 2; therefore, the ranges are from 2 to 8 points, except 2 to 6 in the case of the Environmental dimension. The percentage of the mean score compared to the expected value of assessing indicators also show the lowest value of Environmental sustainability (63.46%). These results imply that the assessors are concerned about Environmental sustainability as it relates to the quality of the input water and the output water. As mentioned above, the water sources for water supply systems in the area are at risk of pollution as a consequence of many factors. In the context of industrialization and modernization in the area, the environmental issue is a great concern, not only for the local authorities, managers, and operators but also for the local people.

However, there is not much difference among the mean values of scores of the dimensions. Specifically, Social dimension was assessed with the highest mean and then followed by Institutional, Financial, and Environmental ones with the values of 6.231; 5.962; 5.769; and 5.077, respectively. These figures imply that the rural water supply systems play an important role in solving related-issues in the area. The two dimensions of Managerial sustainability and Technical sustainability are with the mean values of 3.019 and 2.923, respectively. The standard deviation (SD) between the water supply systems with the largest value is the Financial sustainability index (2.193). This shows that there are systems with financial sustainability; meanwhile, others are not.

The assessment results show a mix of systems in the range from unsustainable to highly sustainable ones. The proportion of the overall results with respect to different categories is shown in Figure 4.

Figure 4 shows that nearly half of the rural water supply system in Ha Nam province was assessed to be highly sustainable (accounting for 44%). In the second rank, 27% of the systems were classified as slightly sustainable. The systems with moderately sustainable and unsustainable occupied approximately 15% and 14%, respectively.

After the sustainability levels of the water systems were measured and determined, an ordered logit regression model was constructed. A descriptive summary of specific variables is given in

Table 7. Descriptive summary of variables.

No	Variable	Observation	Mean	Std. Dev.	Min	Max
1	Model_type (1—Commune People’s Committee, 2—Joint stock cooperative, and 3—Private enterprise)	52	2.06	0.96	1	3
2	Lifespan (years)	52	16.44	5.75	9	27
3	Design_cap (m3/day-night)	52	1427.92	1367.19	50	5300
4	Performance (%)	52	50.9	34.92	0	100
5	Invest_rate (mill. VND/m3)	52	11.86	7.11	2.23	31.31
6	Sustainability	52	2.90	1.12	1	4

Source: Calculated by the authors.

.

Table 7 shows a large difference in the cost of investment per unit of capacity of the water system. This difference reflects the fact that different scales or natural characteristics may have an effect on the investment unit rate. In this study, the data also show that the more recently built the water system, the higher the unit rate of investment.

From the data of the variables described above, an ordered logit model was established using the software STATA 14. The estimate of parameters from the model shows the effect of the examined variables on the sustainability of the water supply systems and is given in

Table 8. Estimated results of the ordered logit model.

Variable	Coefficient	Odds Ratio
Model_type	1.6625 ** (0.8237)	5.2725 ** (4.3430)
Lifespan	0.0733 (0.1299)	1.0760 (0.1398)
Design_cap	0.0018 ** (0.0008)	1.0019 ** (0.0008)
Performance	0.0866 *** (0.0216)	1.0915 *** (0.0235)
Invest_rate	−0.0337 (0.0847)	0.9668 (0.0819)
Constant cut 1	4.4244 (3.6250)	4.4244 (3.6250)
Constant cut 2	8.9218 (3.8534)	8.9218 (3.8534)
Constant cut 3	11.6166 (4.1785)	11.6166 (4.1785)
Observations	52

Note: Standard errors in parentheses *** p < 0.01, ** p < 0.05. Chi2(5) = 76.74, prob > Chi2 = 0.000, Pseudo R² = 0.5801). Source: Calculated by the authors.

.

Table 8 shows that the model has three cut points (because the dependent variable is classified into four ordered categories). The sign and level of significance of the parameters indicate the effect direction of the independent variables on the dependent variable (the sustainability of the water systems); however, they cannot explain in detail for each group of the categorical variable. The Odds ratios indicate how much the dependent variable will change when the independent variables increase by one unit in each category. To explain the effect of independent variables on each category, marginal effects are used.

With the hypothesis H_o that all coefficients of the explanatory variables are equal 0, the LR test follows a Chi-squared distribution with degrees of freedom equal to the number of explanatory variables (=5). In the case of this model with 5 independent variables, the LR Chi-squared value is 76.74. If the hypothesis H_o, then the probability of having a Chi-squared value is equal to or greater than 76.74 (the resulting current value is 0.000 < 0.05, then the model is statistically significant). Therefore, the H_o was rejected, meaning the independent variables can affect the Sustainability variable [43].

Among the variables included in the model, 3 variables were statistically significant, including Model_type (p = 0.044 < 0.05), Design-cap (p = 0.023 < 0.05), and Performance (p = 0.000 < 0.01). The remaining variables were not statistically significant with the p-value > 0.1, including Lifespan (p = 0.573) and Invest_rate (p = 0.691). The coefficients reflect the change of the Log-odds Sustainability of the dependent variable when the independent variable increases by 1 unit, as follows:

-

If Model_type increases by one unit (i.e., the change from Cooperative model to Private enterprise type), the odds of Highly sustainable versus the combined lower category are 5.2725 times greater than if the model_type was Commune People’s Committee or Cooperative model (provided other variables are unchanged). This indicates that the higher the value of the management model, the better the sustainability of the water supply system. The value of the management model was taking the value of (1) Commune People’s Committee model, (2) the cooperative model, and (3) private enterprise; therefore, the private enterprise management model is more effective than the remaining models. The variable Model_type has the highest impact factor of all the model’s variables.

-

If Design_cap increases by one unit (i.e., 1 m³/day.night), the odds of Highly sustainable versus the combined lower category are 1.0019 times greater than if Design_cap did not increase (assuming other variables are constant).

-

If Performance increases by one unit (i.e., 1 percent), the odds of Highly sustainable versus the combined lower category are 1.0915 times greater than if Performance did not increase (given other variables are constant).

Therefore, the results of regression analysis indicate that the management model of water supply systems under the form of private enterprises outperforms and is the most sustainable model compared to the other ones. This finding is similar to the result from a number of previous studies on management models (level of community participation), and system capacity related to the sustainability of water supply systems, such as Cunha Marques [44] and Masduqi et al. [45].

In order to explain the influence of the independent variables on each category of sustainability of the rural water supply systems, it is necessary to calculate the coefficient of its marginal effects [46,47]. The results of estimating marginal effects from the current study are shown in

Table 9. The marginal effect after the logit of different levels of Sustainability.

TT	Independent Variables	Unsustainable	Lowly Sustainable	Moderately Sustainable	Highly Sustainable
1	Model_type	−0.0017 (0.0026)	−0.1289 (0.1023)	−0.2737 (0.176)	0.4044 ** (0.1964)
2	Lifespan	−0.0000 (0.0002)	−0.0056 (0.0099)	0.0121 (0.0229)	0.0178 (0.0318)
3	Design_cap	−0.0001 (0.0000)	−0.0001 ** (0.00007)	−0.0003 (0.0002)	0.0005 ** (0.00023)
4	Performance	−0.0000 (0.0001)	−0.0067 * (0.0038)	0.0142 * (0.0082)	0.0211 *** (0.0061)
5	Invest_rate	0.00035 (0.0001)	0.0026 (0.0068)	0.0055 (0.0141)	−0.0082 (0.0201)

Standard errors in parentheses: *** p < 0.01, ** p < 0.05, * p < 0.1. Source: Calculated by the authors.

. The influence of factors on the sustainability of the water supply system with respect to each category (Unsustainable, Lowly sustainable, Sustainable, Highly sustainable) is significantly different.

Table 9 shows three variables that affect the sustainability of the rural water supply systems in the study region, including the type of management model, the design capacity of the water systems, and the performance of the systems. In these cases, the marginal effects are negative and turn positive at a higher level of sustainability. This is because the marginal effects are technically linear predictors, and they begin below zero and turn over zero. These variables are statistically significant at some levels of the model. The sustainability assessed as ‘Highly sustainable’ is positive with Model_type; meanwhile, it is negative in other cases. More specifically, if Model_type increases by one unit (i.e., the change from Commune People’s Committee model to Cooperative type), the probability of the sustainability moving from Unsustainable to Lowly sustainable increases by 0.17%, from Lowly sustainable to Moderately sustainable by 12.89%, and from Moderately sustainable to Highly sustainable by 22.37% (given other variables are constant). This indicated that the management model has a high impact on the sustainability of the water systems.

This result shows similar findings suggested from studies on the management model of water supply systems due to the different characteristics of each management model, such as financial structure, human capacity, and frequency of system maintenance [21,30,36].

6. Conclusions

This study attempted to examine the effect of management models on the sustainability of rural water supply systems in a rural region of Vietnam. A procedure has been proposed and adopted to determine the sustainability of the water systems and quantify the effect of sustainability management models by using the ordered logit regression method. The investigation on the sustainability of the water supply systems from the study cases showed a wide range from Highly sustainable to Unsustainable. Among the dimensions affecting sustainability in the current context, the Social dimension was ranked as the highest (with a mean score of 6.23), followed by the Institutional and Financial dimensions (with a mean scores of 5.96 and 5.76, respectively). These results differ with the findings of studies from other contexts, such as the economic and technical dimensions from the case of Algeria [19] or the Managerial and Social dimensions in the case of Indonesia [27].

The results from the ordered logit model showed that the sustainability of rural water supply systems from the current cases is significantly affected by a number of factors, including the type of management model, the design capacity of water systems, and the performance of the systems. This finding implies that some forms of intervention in the rural water sector can be taken into account to increase the sustainability of rural water systems in the country.

From this study, a number of policy changes should be promoted to move toward better rural water supply in the study context as follows:

First, it is essential to shift the models of rural water supply systems to the form that increases the role of enterprises and the private sector in investing in and managing water systems. Among the various models in the current study, the enterprise/private management model is expected to be more sustainable than the joint stock cooperative and commune People’s Committee ones.

Second, institutional factors were determined as the most important aspects in the current context, the existence of an appropriate institutional framework that supports the development of rural water supply models as well as guarantees the sustainable management and exploitation of water entities is crucial. This should include transparency in the existing regulations and policies relating to the rural water sector.

Third, the assessment of Environmental factors showed the lowest score in most water systems; therefore, environmental problems embedded with rural water supply are considered critical. This issue has been common in a developing country like Vietnam. Thus, it needs to strengthen the awareness of communities as well as the necessity for the institutions to increase the management of the water environment of Vietnam. This is to ensure that the protection of the environment in general, and the quality of the water environment in particular, is guaranteed.

Although this study achieved some results, it involves some limitations. First, there are many more models of rural water supply management in the context of Vietnam; this study investigated the area with only three existing types of models. A study to investigate a more variety of management models would help to inform a better representative result for the context of Vietnam. Second, in terms of econometric analysis, the sample size of this study only meets the minimum requirement for establishing an econometric model. The relatively small sample size may cause biases in the parameters of the model. Future research with a larger sample size and a more representative of provinces across the country would increase the statistically significant of the results as well as indicate more implications to policymakers. Third, the scores for each level of sustainability given in this study were qualitatively given via consultation of system operators in the field. A quantitative approach with the participation of more stakeholders should be used to avoid some kind of bias.

In short, this study emphasizes the role of management models in affecting the sustainability of rural water supply systems. In the current context, enterprise/private model is considered the most appropriate to guarantee the sustainability of rural water systems. The study provides lessons learnt for Vietnam as a whole and other countries to pursue and promote the sustainability of rural water supply in the future, especially in developing countries.