Why do people save water? A systematic review of household water consumption behaviour in times of water availability uncertainty

As reported in the Sixth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC 2021), continued global warming is projected to increase the severity of droughts. Extreme drought events will become more frequent; hence, society must develop resilience to face this new reality, including enhancing actions to promote water conservation. Such actions are much needed because drought conditions and water conservation are closely intermingled (Echeverría 2020).

The 2030 Agenda, designed by the United Nations, establishes 17 sustainable development goals (SDGs) to direct efforts to achieve a sustainable world. Goal 6 points out the need to ensure the availability and sustainable management of water, which is quite a challenge, since many countries suffer from scarcity, deterioration of quality, and contamination of water bodies, due to the unsustainable management (Wang et al. 2019; Ferasso et al. 2021; Pinto et al. 2021). As water scarcity intensifies in many regions, current interest has shifted from an emphasis on water supply to a more balanced vision considering demand as well through water demand management (Al-Zahrani et al. 2013). For such a shift in water conservation goals to be successful, a better knowledge of the factors that affect domestic water demand is necessary (Villar-Navascués & Pérez-Morales 2018).

Considering this scenario, this review seeks to answer the following question: which factors influence household water consumption in the context of water scarcity? To do so, we collect the main references on residential water consumption and evaluate which factors are more relevant to shaping residential water consumption behaviour in drought contexts, using those factors to organize the literature. Based on an understanding of how such factors are interrelated, we propose a framework that may help future research as well as water demand-side management (DSM) policy design.

Water scarcity is a general state of water insufficiency (Oki & Quiocho 2020). The term ‘water scarcity’ refers to the condition in which demand for water cannot be fully satisfied, generally due to the quality of the water or the impact of water use on supply (Liu et al. 2017; IPCC 2022). Physical water scarcity may be defined as the insufficiency or lack of water itself (Oki & Quiocho 2020), which is usually seen in densely populated arid areas (Rijsberman 2006). There is also an economic type of water scarcity, associated with inadequate infrastructure development and poor management as well as cases of inequitable water allocation and access across economic levels (Oki & Quiocho 2020; IPCC 2022). Finally, water scarcity may be related to the availability of water of acceptable quality, and, in this sense, water scarcity can be reduced through improved wastewater treatment, reductions in pollutant emissions and increasing water reuse within and among sectors (Van Vilet et al. 2017).

Water scarcity and household water consumption have received increasing attention on national public agendas (Inman & Jeffrey 2006), as well as in academia (Liu et al. 2017), for the last two decades. The reason is not only that nearly half of the world's population is currently experiencing severe water scarcity for at least some part of the year due to climatic and non-climatic drivers (IPCC 2022) but also that water shortages are a significant social and economic issue in many countries (Lowe et al. 2014). Understanding water scarcity is important for formulating policies at local, regional, and national scales (Liu et al. 2017).

Al-Zahrani et al. (2013) define DSM as a managerial approach that aims to meet the demand for water through the application of necessary and efficient measures and incentives to achieve fair and effective utilization of water. DSM can be used to reduce water consumption during daily peak use (Beal et al. 2016) and to change long-term water consumption behaviour (Koop et al. 2019). DSM is essential to ensure the availability of water resources (Ibáñez-Rueda et al. 2021).

DSM can be related to economic sectors, such as industry or agriculture, but may also target households, i.e., small groups of people who live together and share living arrangements (Rees et al. 2020). In the domestic sphere, water consumption is largely conditioned by the habits and behaviours of individuals (Ibáñez-Rueda et al. 2021), and in this sphere, DSM campaigns usually involve the co-implementation of several tools, such as technological, financial, legislative, and educational tools (Inman & Jeffrey 2006). DSM programs can be expected to reduce water consumption, but their success depends on the type of approach used (Renwick & Green 2000).

A literature review usually aims to reveal trends, relations, inconsistencies, and gaps in the literature as well as to organize and evaluate work in a particular field (Hahn & Kühnen 2013). We identified four systematic reviews related to this research that only partially overlap. The reviews found during our search were developed in recent years, dating from 2019 to 2021.

A systematic review regarding DSM instruments was carried out by Abu-Bakar et al. (2021), but focused on water consumption measurement, such as ‘micro-components’, ‘end use analysis’, ‘smart meters’, and ‘data-loggers’. Similarly, Roshan & Kumar (2020) reviewed 16 articles and presented an analysis of end-use behaviour that focused on six end uses: toilet flushing, shower/bathtub, dishwashing, indoor taps, laundry, and other/leaks.

Echeverría (2020) analysed 65 articles to understand the social dimensions driving the adoption of water-saving behaviours in urban households, but focused on the variations of water management strategies in different regions of the world. After presenting five key social categories of the analysis reported in the literature – (i) water prices, (ii) water-saving devices, (iii) inflation of self-reported water savings, (iv) intention/behaviour gaps, and (v) weakening of water reductions – the author concluded that strategies to limit water consumption patterns are related to the availability of technological advances, time allocation preferences and the implementation of home appliances. This author also highlighted that in both developing and developed countries, there is behavioural regularity reflecting a gap between intention and effective behaviour in conserving water in households.

Finally, Koop et al. (2019) analysed 58 papers, including empirically oriented studies that investigated how different behaviour influencing tactics (BITs) can stimulate domestic water-conservation behaviour. From these studies, they identified eight key reflective, semi-reflective and automatic BITs that proved to be crucial: knowledge transfer and increasing self-efficacy (considered reflective BITs); social norms, framing, and tailoring (considered semi-reflective BITs); and emotional shortcuts, priming, and nudging (considered automatic BITs).

To comprehend the state-of-the-art in residential water consumption within water scarcity contexts, we conducted a structured systematic literature review. Structured literature reviews mitigate researcher bias and facilitate reproducible outcomes (Pickering & Byrne 2014); furthermore, they offer a collective insight into a given field (Tranfield et al. 2003). For our review, we adopted the following steps, derived from Pickering & Byrne's (2014) and Tranfield et al.’s (2003) procedures: (i) formulating the research question; (ii) defining search criteria, encompassing designated databases, publication time frames, publication types, and search terms; (iii) executing the selection process; and (iv) evaluating the outcomes. The research protocol specifying each of these steps is available as a Supplementary File of this article.

This study aims to overview recent research on factors influencing water consumption behaviour amidst changing global water availability. Thus, our structured literature review was guided by the question: Which factors influence household water consumption in water scarcity contexts?

For our search, we utilized the Scopus and Web of Science databases, chosen for their interdisciplinary coverage and credibility (Mongeon & Paul-Hus 2016). We focused on articles published between 2010 and 2022, considering the pivotal discussions on water management during the Conference of Parties 16 (the CoP16), in 2010, and the increasing awareness of climate impacts on water resources (White et al. 2017). We included only research or review articles in English, using a search string with the following words and terms (and their variant forms): ‘water consumption’; ‘behaviour’; ‘drought’ or ‘scarcity’ or ‘water stress’ or ‘water crisis’; and ‘residential’ or ‘household’ or ‘urban’. This string was applied to titles, abstracts, and/or keywords to retrieve relevant literature.

The selection process commenced with the evaluation of articles' titles and abstracts. Initially, papers not cantered on residential water consumption behaviour or those lacking analysis of specific instances of water scarcity were excluded.

To ensure research reliability, two researchers (i.e., the coauthors) conducted the selection process analysis, adhering to the methodology outlined by Klewitz & Hansen (2014).

For the final stage of the review process, we employed the literature analysis framework recommended by Klewitz & Hansen (2014), which includes a descriptive analysis categorizing papers by year of publication, study type, country of origin, and journal of publication. This descriptive analysis, akin to bibliometric analysis, is particularly useful for mapping scientific trends across disciplines, especially when dealing with extensive, fragmented, or contentious research areas (Aria & Cuccurullo 2017).

Furthermore, for literature evaluation, we conducted content analysis, involving coding followed by interpretation of the coded content (Gaur & Kumar 2018). The focus of this content analysis was to categorize the objective of each analysed article (Navarrete et al. 2020).

The results are divided into two parts: first, we provide a quantitative descriptive (bibliometric) analysis to present an overview of the characteristics and developments in the field of the studied topic. Then, to illustrate the current state of knowledge regarding the influence of water scarcity on residential water consumption behaviour, we present a qualitative thematic analysis of the literature to address the determinants of such behaviour and discuss the main research findings.

Articles regarding water consumption behaviour in the context of droughts are usually published in journals that focus on either water issues or environmental/sustainability issues. There is no journal that concentrates publications on this subject, but ‘Water’, ‘Sustainability’, and ‘Journal of Environmental Management’ were dominant (each with four articles from those included in this systematic review), followed by ‘Water Science and Technology: Water Supply’ (with three articles) and ‘International Journal of Water Resources Development’ and ‘Journal of Cleaner Production’ (with two articles each). The remaining 36 articles were published in 36 different journals.

Most of the studies analysed in this review were based on case studies (single, illustrative, and multiple case studies). Official databases were mainly used to collect information for the research, but some papers were based on surveys (11), and one was even based on focus group interviews. Most of the analyses used quantitative approaches, usually regressions, to identify factors impacting water consumption, but some used multiple methods, systematic reviews, and research based on experiments.

Overall, the literature points to six main factors that shape water consumption (Figure 4): (i) behavioural factors; (ii) water-saving technologies and devices; (iii) water availability context; (iv) awareness creation (through information and education); (v) policy drivers for water saving; and (vi) socioeconomic and demographic factors. There is no trend of a specific topic being studied in one particular year or in a sequence of years, meaning that all six water consumption factors present in the literature were distributed over time during the period of analysis. A relevant fact, however, is that developed and developing countries differ in terms of the most analysed topics: the top three factors that influenced water consumption studied in developed countries were psychological factors, water availability context and awareness, while the top three in developing countries were technology, sociodemographic factors, and policy.

Psychological factors were the main topic in 13 out of the 55 studies from the present review. We considered psychological factors to be those related to perceptions of water rights, environmental threats, social desirability, beliefs, attitudes, and emotions that may impact household practices around water consumption (Radonic 2019; Ibáñez-Rueda et al. 2021).

A few theories were related to psychological factors, but the most cited in the reviewed articles was the theory of planned behaviour (TPB). TPB was proposed by Ajzen (1991) and brought up by some of the analysed articles (Koop et al. 2019; Mankad et al. 2019; Shahangian et al. 2022). According to this theory, people are likely to engage in a particular behaviour if they think their peer group (i.e., people whose opinions they value) expects them to act in such a manner and hold a positive attitude towards that behaviour (Mankad et al. 2019). In this line, peer pressure for water saving might be more effective than government incentives, which is why understanding and being able to shape neighbourhood effects might become a useful lever to facilitate initiatives aimed at promoting community engagement (Wang & Dong 2017).

Peer pressure can also be referred to as social norms, and in recent years, social norm research has emerged in the water-conservation literature because of the key role of social norms in individual attitudes towards water saving (Wolters 2014; Wang & Dong 2017; Ibrahim et al. 2018; Mankad et al. 2019). Personal and social norms are used in the pro-environmental behaviour literature to examine people's sense of moral obligation to perform environmentally correct actions (Mankad et al. 2019), including water saving. The impositions of social norms are important to create a long-standing water-saving culture in communities (Wang & Dong 2017).

Environmental concern is also a strong candidate for influencing water consumption. Nevertheless, environmental concern alone is not a strong predictor of environmental behaviour (Wolters 2014). Previous research has shown that understanding water issues in the community is an important element for solving water-related problems (Wang & Dong 2017); however, varying human judgments – for instance, concerns about water scarcity – are best predicted by different sets of sociocultural factors. This finding supports the value of multifaceted approaches to understanding public perspectives on complex environmental challenges (Larson et al. 2011), such as water scarcity.

Proximity to nature, whether physical or psychological, may also interfere with people's water consumption behaviour (Wolters 2014; Ibáñez-Rueda et al. 2021). Thus, the potential of natural connectedness to change water use patterns should be considered, and interventions that strengthen individuals' relationship with nature by reinforcing their feelings of connectedness could be a good strategy (Ibáñez-Rueda et al. 2021).

One way to reduce water consumption is through the adoption of technologies and devices that aim to save water. In this review, one-fifth (11 out of 55) of the analysed articles had ‘water-saving technologies and devices’ as the main topic, and another 12 included water-saving technology in their analysis, but not as the main topic. Corroborating Edirisinghe & Pathirana's (2021) conclusion that there has been increasing interest in this matter in the literature, in our review, two-thirds of the articles that focused on technological factors were published from 2019 on.

Understanding the determinants of the adoption of water-efficient devices is an important step towards the development and incorporation of technological policies, particularly policies seeking to reduce water consumption. That is why such an understanding is commonly addressed in the literature (see Martínez-Espiñeira & García-Valiñas 2013; Ramsey et al. 2017; Koop et al. 2019; Radonic 2019; Rule et al. 2021), particularly when related to some of the most used technologies, such as dual-flush toilets, flow-control devices, water-saving dishwashers, or washing machines with lower water consumption and devices and methods for grey-water collection and use (Martínez-Espiñeira & García-Valiñas 2013; Abu-Bakar et al. 2021).

The acquisition and use of water-saving devices depend on socioeconomic and demographic factors, such as education, financial capacity, and age (Martínez-Espiñeira & García-Valiñas 2013; Rule et al. 2021). Acquisition may be intensified by policies that encourage the adoption of water-saving devices, such as subsidies (Martínez-Espiñeira & García-Valiñas 2013), and by water availability context (Stone & Johnson 2022). However, it is worth keeping in mind that water-conservation behaviours that require little effort are most common, while conservation behaviours that require additional effort or financial investment are less popular (Ramsey et al. 2017). That is why policies should encourage consumers to purchase relatively cheap and easy-to-install water-saving mechanisms (March et al. 2015). Finally, the incorporation of water-saving technologies may also be influenced by the perceived quality of water derived from the use of such devices. For instance, awareness campaigns should propagate information to reduce risk and perceptions of water produced by the desalination process as low quality (Villar-Navascués & Fragkou 2021).

Analysis of the available DSM technology reveals that the existing methods and equipment are available not only for end users but also for water utilities. Understanding consumption behaviour can lead to successful categorization based on the recognition of similarities in consumption patterns among consumers (Ioannou et al. 2021) and help forecast future trends, which is crucial for adequate demand management (de Oliveira et al. 2020). In the same line, applying data analytics and machine learning to data gathered from digital water meters may be a useful tool for gaining a better understanding of consumer behaviours, habits, and routines and, consequently, for creating more effective water conservation and demand management strategies (Rahim et al. 2021).

A field that is not yet commonly explored in the literature is Internet of Things (IoT) based methods, as only one of the 33 articles that explored water-saving technologies focused on this subject. The importance of studying such methods is related to the possibility of monitoring household appliances via wireless systems, collecting data in a central database and using those data to intervene in the consumption behaviour of households (Edirisinghe & Pathirana 2021).

Interestingly, in this review, a considerable number of articles that examined water-saving technologies focused on rainwater harvesting systems (Moy 2012; Hunt & Rogers 2014; Ramsey et al. 2017; Radonic 2019), and because of these evaluations, one might conclude that the use of water from such sources may vary. For instance, in Australia, Asia, and Europe, rainwater collection systems are increasingly connected to indoor non-potable end uses, such as toilet flushing and laundry (Moy 2012; Hunt & Rogers 2014; Ramsey et al. 2017), while in the United States, rainwater tanks are promoted and used primarily for outdoor irrigation (Radonic 2019).

Regarding rainwater harvesting, it is important to highlight that the installation of such systems does not necessarily result in a reduction in potable water use (Ramsey et al. 2017; Radonic 2019). Additionally, the saving potential of rainwater tanks will not be realized and help reduce pressure on water sources if the sociocultural context is not recognized as the most influential determinant of consumption (Radonic 2019).

Ten out of the 55 papers reviewed focused on public awareness as a factor in creating water-saving behaviours. The ways of creating awareness most explored in the literature were information dissemination (Hodges et al. 2020; Hunt & Shahab 2021; Ong & Aral 2021; Ostad-Ali-Askar et al. 2018), educational campaigns (Lowe et al. 2014; Quesnel & Ajami 2017; Ramsey et al. 2017), and benchmarks and feedback (Hunt & Rogers 2014; Hoppins & Meshes 2022).

Although information and education were the most studied methods and are likely to be necessary, they were not sufficient components of governmental programs for behavioural change (Martínez-Espiñeira & García-Valiñas 2013; Hunt & Shahab 2021). Rather, a combination of factors is needed to promote water-saving behaviour, such as restrictions and subsidies (March et al. 2015), available technologies (Hunt & Shahab 2021; Shahangian et al. 2022), and sociodemographic and economic characteristics of the target population (Hunt & Shahab 2021).

Educational campaigns that teach easy ways to conserve water may increase the feeling of self-efficacy, leading to recommendations for how the adoption of certain behaviours can make a difference (Lowe et al. 2014; Ramsey et al. 2017). Communication and knowledge create awareness and induce individuals to act more responsibly by developing the personal need to act correctly (Hunt & Shahab 2021).

There are different forms of awareness creation. One is through news media, which was proven to be an efficient mechanism for reducing residential water use at the fastest possible rate (Quesnel & Ajami 2017) as well as a potential way to encourage water-conservation behaviours by demonstrating that the government had invested in improving water security (Ramsey et al. 2017). Information dissemination through social media was another form with the potential to be an effective approach to change the use behaviours of household water consumers. The literature suggests that such communication is valued by people because it reduces their knowledge deficit and increases conservation behaviours (Lowe et al. 2014).

Awareness campaigns should regularly revise their scope and their channels of diffusion to guarantee their efficiency and to include water-saving innovations (March et al. 2015; Wang & Dong 2017). Water awareness campaigns could benefit from more accurate knowledge of the habits of water use in targeted areas (March et al. 2015). Such knowledge includes information on both the success of past awareness campaign messages on different users and the evolution of water use habits.

For many years, these were the most explored factors in the literature since they have been shown to be some of the most significant in shaping household water consumption behaviour (Damodaran 2011; March et al. 2012). Nevertheless, in more recent years, studies have shown that these factors alone are inconsistent in predicting environmental behaviours; thus, they have been used mainly to segment information, such as creating clusters (i.e., gathering households/inhabitants with similar characteristics) and investigating similar consumption patterns within such clusters (Wolters 2014; Ibrahim et al. 2018, 2021; Villar-Navascués & Pérez-Morales 2018; Radonic 2019; Sousa & Fouto 2019; Ito et al. 2021; Kumar et al. 2021; Sousa et al. 2022). In the present literature review, socioeconomic and demographic factors were the main topics in only 8 of the 55 articles; however, variables associated with these factors were present in more than half of the studies read for this review (33 out of 55).

Socioeconomic and demographic factors are commonly used in models in studies of policies and other governmental actions towards water saving (Mini et al. 2014; Yuan et al. 2014; Brelsford & Abbott 2017; Biswas & Gangwar 2021). This is mostly because understanding different behaviours among diverse sociodemographic groups provides opportunities for targeted interventions to enhance water conservation (Wolters 2014; Wang & Dong 2017).

March et al. (2012) and Rees et al. (2020) included a social factor that was not often considered in the articles analysed but seems to be relevant: immigration. Consumption behaviours for immigrants seem to be in line with characteristics that are usually explained by three features from economic, sociodemographic, and cultural dimensions: low-income, high-density habitation, and religious practices (March et al. 2012; Ahmad & Daura 2019; Rees et al. 2020). However, some studies showed that cultural domains explain environmental perspectives better than demographic factors, influencing effective and cognitive judgments about water issues (Larson et al. 2011).

The differences in economic characteristics among countries and their responses to water shortages are also important but have not been extensively explored in the literature. There are differences in the trends of water consumption between developed and developing nations, thereby implying a need for separate prospects of end-use results in these countries (Abu-Bakar et al. 2021). For instance, some authors have identified a declining trend in household water consumption in developed counties (March et al. 2012; Saurí 2020), while in developing countries, especially those going through the urbanization process, per capita water consumption has increased in recent years (Damodaran 2011; Ibrahim et al. 2021; Ito et al. 2021). Additionally, there are some particularities in developing countries, such as many people with no access to piped water, people living in slums (Ibrahim et al. 2021) and high levels of water lost during distribution, that could interfere with reduced water use and water consumption. The predominance of studies focusing on developed countries, evidenced in our descriptive analysis, has previously been criticized by Koop et al. (2019), who claimed that although most studies in the literature have been undertaken in developed countries experiencing water stress, water scarcity is clearly a problematic issue in many other parts of the world.

The context of water availability was the main object of 7 out of the 55 papers reviewed, a low number considering that scarcity events are becoming a more frequent problem. This evidence supports Vallès-Casas et al.’s (2017) conclusions that the effects of contingent situations, such as drought episodes, remain less explored in the literature than other factors that influence household water consumption.

Although the relation between droughts and water-saving attitudes was not a ‘hot topic’, studies that focused on understanding this subject evidenced that the water scarcity context induced people to incorporate water-saving habits (Bernardo et al. 2015; Beal et al. 2016; Wang & Dong 2017; Sousa & Fouto 2019; Stone & Johnson 2022) Developing research on this topic is crucial to help understand the relation of drought and other factors that influence such habits and to identify the adopted behaviours to gather evidence that can be used in water-saving actions and policies.

The literature indicates that there are preferred actions and behaviours regarding reducing water use in a water scarcity context (Lindsay & Supski 2017; Wang & Dong 2017), such as outdoor water uses, including swimming pools and watering gardens. That is why, as Wang & Dong (2017) concluded, the outdoors is becoming an important place for people to implement water-saving practices.

The experience of drought was reported as a factor that changed consumer behaviour even after a drought was over (Beal et al. 2016; Bernardo et al. 2015; Wang & Dong 2017; Sousa et al. 2022; Stone & Johnson 2022). This means that during water scarcity periods, people incorporate new habits that remain even when the dry periods end. Nevertheless, behaviour and attitudes are not permanent and can change based on new information and experiences (Mankad et al. 2019).

Inhabitants of locations that more frequently struggle with water scarcity tended to be more willing to conserve water (Rule et al. 2021). Although it seems to be common sense that water conservation is important and is everyone's responsibility, water-conservation attitudes, and behaviours differ significantly in different locations, depending on the water context: inhabitants of drought-prone regions perceive more pressure to conserve water and try to conserve water wherever they can.

Finally, regarding other factors that, with water scarcity, shape water use behaviour, Shahangian et al. (2022) identified a perception of a link between water scarcity and health and concluded that perceived severity derived from water scarcity-related health problems.

Policy-driven factors were the main topic in only 6 out of the 55 articles reviewed; however, they were also present in the other 19 studies. Recently, both water suppliers and public authorities have focused on demand-side water policies as a preferred alternative to supply-side initiatives (Martínez-Espiñeira & García-Valiñas 2013). The first type includes public outreach, free waterwise landscaping consultation, tax rebates for waterwise landscaping, subsidized technologies for recycling wastewater, increasing awareness, and economic incentives, such as subsidies, fines for wasting water, and tariff changes (Wang & Dong 2017; Sousa & Fouto 2019). The second type includes reductions in water supply duration, an action that can have adverse effects, such as the creation of a scarcity mindset that leads people to be less likely to engage in water-conservation behaviours (Ramsey et al. 2017). Additionally, reductions in water consumption because of water supply restrictions tend to decrease over time (Tsuda et al. 2014).

Policies based on economic instruments have been shown to be effective in some cases reported in the literature (Lucas & Cordery 2019; Stone & Johnson 2022). On the other hand, other authors have claimed that economic incentives may have adverse results, such as problems of equity and fairness in the social effectiveness of economic measures (March et al. 2015) and may be poorly chosen by governments that are concerned about electoral cycles (Grover & Lucinda 2021).

The literature has shown that mandatory and punitive policies may be more effective than policies such as voluntary watering restrictions. This finding has been reported for both indoor and outdoor water consumption reduction (Mini et al. 2014; Wang & Dong 2017; Stone & Johnson 2022). Nevertheless, programs aiming to encourage households to take these voluntary actions may succeed, as mentioned before, depending on perceived susceptibility and perceived severity.

Socioeconomic and demographic factors are commonly addressed in studies focusing on water-saving policies (Martínez-Espiñeira & García-Valiñas 2013; Wang & Dong 2017). They are important variables since understanding the differences in behaviour among different sociodemographic groups allows targeted interventions (Wang & Dong 2017). More commonly, the literature reports that higher-income families are more willing to support costly water-saving policies (Wang & Dong 2017) and show larger reductions, probably because there is a greater margin for reducing consumption since they may use water for less essential activities (Sousa & Fouto 2019). Additionally, low-income people who live in slums commonly do not have access to piped water and, therefore, may not respond to economic incentives (Ibrahim et al. 2021) or even other kinds of policies.

Brelsford & Abbott (2017) pointed out a difficulty in distinguishing policy-driven changes from changes driven by broader technological or demographic trends; they commented that since governments often implement different water-focused policies simultaneously, and since such policies usually emerge during periods of social, technological, and demographic change, it is difficult to tell them apart. Additionally, DSM policies work better when decentralized (Abu-Bakar et al. 2021), meaning that the participation of different stakeholders, such as governments and households, regulators, and firms, is rather relevant. Understanding the dynamics of water-conservation behaviours and consumption is helpful when formulating such policies (Ramsey et al. 2017).

Psychological factors are at the centre of the framework, not only because they are the most studied topic in the studies in this review but also because they are directly or indirectly influenced by other factors. They represent behavioural determinants, which are those that better explain, based on the evidence from the articles reviewed, the actions and behaviours derived from stimuli for water saving.

The adoption of water-saving technologies is related to behavioural factors, as their use depends on the perception of the individual that such technology will contribute to conserving resources and will not impact the well-being of the individual (Mankad et al. 2019; Edirisinghe & Pathirana 2021).

The literature also shows that the water availability context influences people to incorporate water-saving behaviour (Wang & Dong 2017; Sousa et al. 2022; Stone & Johnson 2022). This happens because people feel the seriousness of the situation and generally understand that they are part of the problem as well as part of the solution (Gilbertson et al. 2011).

Although a limited number of studies focused specifically on awareness, including information and educational campaigns, we may infer from the reviewed articles that this subject is quite relevant for DSM. That is why, in our proposed framework, awareness impacts technology, water context, psychological factors, and policies. First, information is crucial for people to understand the benefits of implementing new technologies and for overcoming preconceptions (Mankad et al. 2019; Shahangian et al. 2022). Second, awareness of the severity of droughts and the importance of action to overcome the situation can influence people's perception of the need to conserve (Gilbertson et al. 2011). Third, awareness induces individuals to act more responsibly by developing the personal need to act correctly (Lowe et al. 2014; Ramsey et al. 2017; Gómez-Llanos et al. 2020). Finally, information and education are needed to guarantee the efficiency of policies and to include innovations (Wang & Dong 2017).

Socioeconomic and demographic factors, while not the most extensively analysed topics, are prevalent as variables in nearly half of the papers reviewed for this study. Although they may lack consistency as standalone predictors of environmental behaviours (Wolters 2014), they frequently contribute to understanding various aspects, such as the adoption of technologies (Martínez-Espiñeira & García-Valiñas 2013; Ramsey et al. 2017) and the acceptance of policies (Martínez-Espiñeira & García-Valiñas 2013; Brelsford & Abbott 2017; Wang & Dong 2017; Sousa & Fouto 2019).

There is a large body of studies on water-saving policies since demand-side water policies have been used recently by many water suppliers and public authorities (Martínez-Espiñeira & García-Valiñas 2013) and an understanding of the dynamics of water-conservation behaviours is helpful for the design of such policies (Ramsey et al. 2017). Nevertheless, water consumption reduction policies have some limitations. Apolitical approaches may be preferable for some populations (Mankad et al. 2019), considering that water scarcity is an issue that will also affect future generations and thus must be addressed throughout future years, which may not happen if policies are determined by electoral cycles (Grover & Lucinda 2021).

It is important to explore differences in the DSM of the literature from the country-of-origin perspective. Some authors studying developed countries identified a decreasing trend in water consumption (Tsuda et al. 2014; Brelsford & Abbott 2017; Saurí 2020), while there is a low number of studies in developing countries, where scarcity problems are becoming an increasing issue, and this is a barrier to achieving the United Nation's Sustainable Development Goal (SDG) 6. The management of urban water demand has become urgent in urban centres in developing countries, where there are obsolete supply systems and poor urban planning that is not sensitive to water resources (de Oliveira et al. 2020; Biswas & Gangwar 2021). It is necessary to encourage and conduct studies in these needy regions to provide guidelines for and insights into how to face the problem.

To comprehend household water consumption behaviours, we must consider a set of factors that this research aims to clarify: psychological factors, available technologies, awareness, sociodemographic and economic features, water availability context, and applied policies. The relation of these factors is identified in this review by a framework that aims to support future research as well as policy design. It may not be a wise strategy to analyse each topic separately, as their relation is what determines water consumption behaviour. Overlooking the connections among the factors may result in poor research as well as inefficient policies. However, no research or policy description was found in the literature, including all six factors in their scope.

Overall, we conclude that all the factors analysed somehow influence people's perceptions, which is why psychological factors are the main determinants of water consumption behaviour. The water availability context, existing technologies and policies implemented help build these perceptions, and all these factors depend on socioeconomic and demographic characteristics as well as on people's awareness of the water context and the consequences of their own actions for water availability. To rightly shape people's water consumption behaviour, it is necessary to analyse those determinants and create a policy that fits properly. This is a permanent exercise that requires constant evaluation of possible changes in such factors as well as communication.

The literature still lacks studies that investigate water consumption behaviour in developing countries, where water scarcity is a significant concern. This review highlights the need to address this issue, as understanding the nuanced variations of each factor across different locations will facilitate the development of more effective policies and the attainment of SDG 6. Collaboration between developed and developing countries is essential for achieving these objectives.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.