Comparative economic analysis of urban sanitation interventions in low- and middle-income countries: a systematic review

At present, cost-effectiveness data for sanitation options in LMICs are limited to order-of-magnitude estimates (Hutton 2013). Such data are insufficient for making municipal-level decisions about urban sanitation standards (Hutton & Chase 2016). Inclusive selection of least-cost sanitation options and standards must account for the heterogeneity of cost determinants in LMICs. These include population density, topography, existing service models, sociodemographic profiles, and household practices and behaviours (Daudey 2018). Existing data only partially reflect user experiences (Hutton et al. 2014). Evidence-based decisions often overlook unknown costs associated with informal service provision in cities in LMICs. This includes services not provided by utilities, especially self-supplied systems (Danert & Hutton 2020). This limitation leads to fragmented service outcomes and an uneven distribution of investment benefits (Lawhon et al. 2017; Lüthi et al. 2020). Such outcomes are challenging to characterise (Hutton & Chase 2016; van Welie et al. 2018).

Current best practices use a heuristic model to simplify comparing the urban sanitation options at the municipal level (Sainati et al. 2020). However, the assumptions facilitating comparisons in cities in LMICs are limited. These include (i) using existing service provision models to characterise infrastructure configurations, (ii) assuming the outcomes of these models are acceptable and equivalent, (iii) that lifecycle costs can be reported in a standardised format, and (iv) that equitable decisions can arise from a single cost perspective. Such assumptions are limited in their capacity to account for the complexities within urban sanitation contexts in LMICs. These assumptions aligned with a generalised comparison of the feasibility of existing alternatives (Sainati et al. 2020), rather than fully embodying citywide inclusive sanitation (CWIS) principles (Lüthi et al. 2020).

Complementary approaches are needed to address the complex urban contexts in LMICs, where diverse service provision models coexist (Lawhon et al. 2017; van Welie et al. 2018). Deliberating trade-offs between multiple competing priorities for CWIS interventions is also necessary in these situations (Lüthi et al. 2020). It is important to clarify how heuristic rules of thumb about selecting sanitation options are integrated with detailed analytical decision support to address issues in complex, evolving local systems (Lawhon et al. 2017; van Welie et al. 2018). Such integration is vital for comparative economic analysis to foster a transition to more inclusive outcomes (Sainati et al. 2020). Therefore, it is important to discuss how effectively comparative economic analysis studies embody CWIS principles.

Recently, evaluative criteria aimed at enhancing the quality of least-cost economic analyses of CWIS interventions were devised from a scoping review (Ross et al. under review). This paper applies these criteria as a charting tool within a systematic review (Khalil & Tricco 2022). It appraises the methods used to design, implement, analyse, and interpret comparative economic analysis studies for urban sanitation configurations within LMICs over the past 20 years. The insights from this review will shape fit-for-purpose methodologies that complement more generalised comparisons and work towards achieving CWIS principles. Specifically, the review emphasises the need to enable diverse mixtures of interventions at different scales to be compared, represent the complexity of the existing context as a basis for comparison, enable both tangible and intangible determinants of lifecycle costs and benefits to inform decisions, and accommodate diverse cost perspectives within decisions, particularly for those with marginal access to sanitation services (Lüthi et al. 2020).

This paper adheres to PRISMA guidelines (Page et al. 2021), as outlined in Figure 1.

The review builds on previous work that examined the costs of urban sanitation interventions in LMICs (Daudey 2018). This earlier work highlighted the scarcity of relevant lifecycle cost data with equivalent objectives, timeframes, and scales of analysis (Whittington et al. 2012). The eligibility criteria are structured using the participant, comparator, intervention, outcome (PICO) format (Khalil & Tricco 2022). We defined the participant as the economic analysis methodologies and the intervention as comparing cost-effectiveness data for urban sanitation options in cities in LMICs. The comparator was the evaluative criteria from Ross et al. (under review), and the outcome focused on how well the methodologies embodied CWIS principles.

We defined search terms to identify and expand on comparative cost analysis studies identified by Daudey (2018) that aligned with our criteria. We searched the Scopus database, using combinations of the search terms ‘urban’, ‘sanitation’, ‘cost*’, ‘economic’, ‘financial’, ‘appraisal’, and ‘method*’ to find English-language literature from 2003 to 2022. Broader terms resulted in many irrelevant results, so we refined our terms to ‘sanitation’, ‘cost*’, and ‘method*’, yielding 1,365 records.

We scanned the titles and abstracts of 1,365 records, yielding 200 results. Of these, 17 were inaccessible or not in English. After reviewing 183 reports, 17 were found to have met the criteria. Studies were excluded for several reasons. Some focused on high-income countries or rural contexts (n = 60). Others had objectives not aligned with sanitation outcomes (n = 38) or lacked detailed economic analyses (n = 68). We included 19 additional studies identified via reference and citation lists of the 183 retrieved studies.

We conducted a detailed review of each study (see Table 1) using evaluative criteria from a previous scoping review (Ross et al. under review). Refer to Table 2 and Supplementary Table S1 for details. We ranked criteria responses on a four-point scale: fully applied, mostly applied, partially applied, and not present, considering the study context, methodology, and supplementary material without seeking additional author insights. No attempt was made to weigh each criterion or evaluate the overall quality of each study retrospectively.

Table 2

Abridged evaluative criteria (Ross et al. under review) and analysis of the included literature

Note: These criteria are presented in full in Ross et al. (under review) and Supplementary Table S1.

View Large

Our search strategy and screening method identified 36 diverse and comprehensive ranges of economic analysis methodologies. These papers encompass 22 different economic appraisal methodologies for urban sanitation interventions. They span 44 cities across 23 countries, covering developments over the past 20 years.

In this section, we analyse the extent to which each study's methodology responds to each evaluative criterion for CWIS interventions.

Criterion 1.1 assesses whether the methodology balances the selection of the most economically efficient option with the equitable distribution of financial costs and benefits of that option. Table 2 demonstrates that three studies partially apply this criterion. These methodologies adjust each option's net financial costs to partially account for their net economic costs. For example, Manga et al. (2020) calculated the economic value of options using shadow pricing for capital, land, and water. Prihandrijanti et al. (2008) and Ross et al. (2016) considered averted health costs, i.e. costs no longer incurred because of a sanitation intervention, in their analysis. Shi et al. (2018) traded off the financial value of different interventions against a measure of their environmental impact. However, these studies are limited as they only consider how costs are perceived from a single institutional perspective, not how they are distributed.

Studies that mostly apply this criterion analyse the economic efficiency of each option and how each option distributes financial costs and benefits to different parties. For example, Meddings et al. (2004) and Schuen & Parkinson (2009) considered the benefits of reducing child mortality and the net economic per capita costs of providing sanitation services in Africa, respectively. They also analysed how the financial costs of each option are distributed to a project and household perspective. Similarly, Delaire et al. (2020) compared the economic efficiency by which different options may achieve adequate citywide, safely managed urban sanitation by 2030. Then, they used a willingness-to-pay survey to analyse how users and government subsidies may cover the financial costs of each option.

Studies that fully apply this criterion analyse how the financial costs of an economically preferred option are perceived by subgroups, such as women, girls, or those living in informal settlements. For example, the Economics of Sanitation Initiative (ESI) in Southeast Asia (Hutton et al. 2014) comprehensively measured the economic efficiency of different options and then determined how the financial costs of these interventions are distributed to households from different income quintiles. Willetts et al. (2010, 2013) demonstrated an alternative approach to fully applying this criterion by integrating a financial cost–benefit analysis for different options into a deliberative economic appraisal. In this case, options are prioritised based on broader economic criteria, including financial and non-financial costs and benefits. Both of these studies respond effectively to CWIS principles (Lüthi et al. 2020) by explicitly accounting for the distribution of costs to marginal groups.

Criterion 1.2 complements Criterion 1.1 and assesses whether a study adopts a societal cost perspective that includes all relevant costs, irrespective of who incurs them. This perspective helps one to consider whether the measure of economic efficiency used to prioritise options is comprehensive, which is key to defining a least-cost-to-society ranking. An example of a study that fully applies this criterion is Hutton et al. (2014). They account for the tangible and intangible economic costs and benefits of various options, from multiple financial cost and subgroup perspectives. However, although the net costs were derived from municipal data, they aimed for national comparisons, making the data less context-specific than in other studies.

Studies that mostly apply the criterion limit the scope of a societal cost perspective. The scope may be limited to particular urban areas. For example, Ross et al. (2016) and Delaire et al. (2020) compared the economic efficiency of options for a ‘slum’ and ‘low-income’ context, respectively. However, this analysis is also limited to this cost perspective, so financial trade-offs with other parties in an urban population cannot be considered. An alternative method limits the societal cost perspective by focusing solely on financial efficiency. This approach assumes that economic outcomes for each option are equivalent. For example, Peal & Georges (2013) and Willetts et al. (2010) compared hybrid mixtures of sanitation interventions aimed at providing services to all cost perspectives within a specific population. These approaches required detailed forecasting of each option's financial costs and benefits to test assumptions over the time horizon of each study. Another method of applying this assumption is to divide an urban area into different service regimes (van Welie et al. 2018) or characteristic service provision models, thus enabling different cost perspectives to be disaggregated from a broader societal cost lens. Proponents assume that a service regime that exists in the same geographical area will have the same economic determinants of costs justifying a financial comparison. Both Dodane et al. (2012) and McConville et al. (2019) applied this approach. However, in both examples, it is unlikely that costs are spatially or temporally homogenous across the study context. This variability limits how meaningful a financial comparison is in practice. This limitation persists even when McConville et al. (2019) attempted to optimise the comparison by scaling options to full capacity. Cost determinants in this case are unlikely to be consistent even within the same service regime. Carrard et al. (2021) further demonstrated how even the boundaries of a single service configuration are challenging to define in terms of geography or population.

Studies that best apply Criterion 1.2 are those that balance the inclusion of all costs to the society and how meaningfully they can be disaggregated and analysed (Lüthi et al. 2020). The examples above illustrate how challenging accounting for heterogeneity within the context of studies of CWIS options can be.

Criterion 2.1 evaluates whether urban sanitation options are compared against meaningful and realistic without-project standards. These standards represent cost-effective situations that would likely occur without any intervention, according to the context of the study. To facilitate compelling decisions, options should be defined in addition to this standard (Willetts et al. 2010). Each option should be mutually exclusive, meaning selecting one precludes the selection of others. Table 2 indicates that no studies fully apply Criterion 2.1. This result reflects the challenges discussed in response to Criterion 1.2 of framing an objective and meaningful basis of comparison.

To illustrate these challenges, we consider Willetts et al. (2010). They mostly apply this criterion by systematically defining a reference case using an end-use model (see Criterion 7.1) (Retamal et al. 2011). They compare costed sanitation service provision scenarios based on detailed concept designs aligned with local policies for new greenfield developments. However, they neglected to include the cost perspectives of existing urban users. This omission makes it less certain whether a least-cost without-project standard (see Criterion 1.2) was fully defined. The local status quo likely included a mix of informal service provision, as well as plans for developing conventional centralised services. This definition is crucial because it affects the inclusivity and relevance of the options compared in the analysis. When faced with the challenges of collecting data to represent a least-cost without-project standard (see Criterion 2.1), many studies define a do-nothing scenario. This assumes that the existing context remains the same over time. The difference in these definitions is nuanced but significant. Defining a do-nothing scenario excludes the opportunity costs of forgoing existing practices from consideration.

Other studies that mostly apply Criterion 2.1 comprise various attempts to define a meaningful without-project context. For instance, Kennedy-Walker et al. (2016) compared various modelled scenarios of faecal sludge collection and transportation with results from an existing pilot study. Dahiya et al. (2006) and Hutton et al. (2014) compared iterative improvements in sanitation service levels against a sanitation service ladder. Delaire et al. (2020) defined investments that would have been made by households and institutions without a project, based on results from a willingness-to-pay survey. Furlong et al. (2020) attempted to contextualise financial flow models validated in a comparable city. At the same time, Carrard et al. (2021) defined a without-project case without comparing it to alternative interventions. Any approach may introduce subjective biases, making a critical evaluation of whether a meaningful basis of comparison has been selected necessary (see Criterion 9.3).

Criterion 2.2 assesses whether the system boundary of comparative economic analysis is clearly defined and consistently applied (Willetts et al. 2010). This idea aligns with the standard, generalisable system boundary proposed by Sainati et al. (2020) for comparing unit costs across different urban sanitation contexts. This system boundary specifies the cost perspectives (Criterion 1.2), lifecycle cost components (Criterion 6.1), and components of a sanitation service chain (Criterion 7.1) relevant to analysis.

Some studies that fully apply Criterion 2.2 such as Peal & Georges (2013), Delaire et al. (2020), and Furlong et al. (2020) adopt comprehensive system boundaries. Others like Kennedy-Walker et al. (2016) may adopt more discrete boundaries specific to a context, such as the collection and treatment of faecal waste for low-income users. Carrard et al. (2021) demonstrated that system maps are valuable for visually depicting the costs and benefits within a system boundary, as well as external factors that might influence a study's results.

Criterion 3.1 evaluates a study's adherence to three required elements of a comparative economic appraisal. These include conducting a cost-effectiveness analysis, optimising each alternative towards a defined service outcome, and engaging end users in the final cost–benefit determination. All studies meeting the review's eligibility criteria fulfil the first two requirements of Criterion 3.1. Therefore, our analysis focuses on how effectively studies engage end users in decision-making processes. Based on this criterion, only the study by Dahiya et al. (2006) fully applied Criterion 3.1. They adopted an informed choice approach, enabling service users to trade off perceived benefits of various household containment options against their financial cost.

Other studies mostly apply Criterion 3.1 by demonstrating methods that engage institutions in a decision using approaches that may also be used to engage service users. For example, Furlong et al. (2020) conducted workshops with local service providers to contextualise how different financial flow models may accurately represent the study context. Willetts et al. (2013) engaged local authorities in deliberating a least-cost option. Delaire et al. (2020) indirectly engaged low-income households in a least-cost determination via a willingness-to-pay survey. Studies fully meeting this criterion should consider how these approaches engage service users in decision-making. Such engagement builds demand for safely managed sanitation, aligning with CWIS principles (see Criterion 10.3).

Criterion 4.1 assesses how specific data sources deriving cost-effectiveness measures, to a context, focus on the relevance of input variables, models, and assumptions. Table 2 indicates that the extent to which different studies describe caveats associated with their data sources varies.

Criterion 4.2 assesses how meaningful cost-effectiveness comparisons across diverse urban sanitation contexts are. In general, urban sanitation interventions in LMICs feature heterogeneous cost determinants, making direct comparisons difficult to justify (Carrard et al. 2021). For instance, Sainati et al. (2020) recommended using generalised estimates with a minimal error of ±25% for broad comparisons. Standardised comparisons, like those in case–control studies, pose challenges for urban sanitation. They require specifying a single cost-effectiveness measure that encompasses all relevant costs and benefits (Ross et al. 2022). Table 2 indicates that no stud fully applies this criterion. However, several acknowledge the limitations of these comparisons (Meddings et al. 2004; Hutton et al. 2014; Carrard et al. 2021). Hutton et al. (2014) suggested using generalised comparisons mainly in broad national contexts. In contrast, the heterogeneous contexts of cities in LMICs demand more detailed analyses.

Criterion 5.1 assesses if studies use standardised unit costs, specifically total annualised cost per household (TACH) or per capita (TACC), to compare sanitation options. This method is advocated by Sainati et al. (2020). These units facilitate an incremental comparison of the lifecycle costs of relevant options in addition to a without-project case (see Criterion 2.1). In this case, instead of assuming a linear change, the net present value of each option varies as households or individual users gain access to water-related services (see Criterion 6.1). This approach enables more meaningful comparisons of diverse urban sanitation interventions, from capital-intensive infrastructure projects to lower-cost, intermediate options with more intermediate service outcomes. Incremental analysis also aids in evaluating demand-side management options and the potential for avoided costs from delaying infrastructure-led options.

Studies that fully apply Criterion 5.1 include Kennedy-Walker et al. (2016), who analysed different faecal sludge collection and transport options for urban poor communities. Manga et al. (2020) compared simplified sewer configurations against onsite options. In addition, Retamal et al. (2011) modelled the financial performance of various hybrid sanitation interventions. These studies exemplify the application of average incremental costing (AIC) (see Criterion 6.1) for projecting detailed lifecycle costs to inform future service provision models.

Criterion 5.2 assesses whether benefit valuation methods for economic efficiency are clearly defined and appropriate for the study context. Benefit valuation for urban sanitation can be contentious due to the potential for bias by only partially including all relevant outcomes of an intervention. The ESI methodology (Hutton et al. 2014) is an example that comprehensively applies Criterion 5.2 via the range of benefits it attempts to quantify using rigorous methods. Typically, benefit valuation methods are most effective in homogeneous settings where they can be applied consistently to yield reproducible results. In heterogeneous urban contexts in LMICs, justifying benefit valuation methods is more difficult, as demonstrated in Table 2, where few studies attempt this. For example, Meddings et al. (2004) conducted a case–control study, Prihandrijanti et al. (2008) estimated averted health expenditures, Delaire et al. (2020) utilised a willingness-to-pay survey, and Schuen & Parkinson (2009) assumed a generic health benefit for sub-Saharan Africa.

Criterion 5.3 evaluates whether studies distinguish between intangible costs and benefits and economic efficiency measures in reported results. Urban sanitation systems often include diverse intangible benefits that are difficult to define and measure. Studies that fully apply Criterion 5.3 systematically differentiate between factors that can and cannot be quantified to support clearer decisions. For instance, Hutton et al. (2014) effectively categorised benefits as valued, quantified, and tangible. Another example is Willetts et al. (2010, 2013), who embedded a quantitative measure of cost-effectiveness within a deliberative sustainability assessment of sanitation options conducted with local authorities. The latter approach further exemplified Criterion 5.3 by transparently aligning their approach for defining and handling externalities with global sustainability frameworks.

Criterion 6.1 assesses whether AIC has been applied to structure lifecycle costs within a study. Table 2 shows that Kennedy-Walker et al. (2016), Manga et al. (2020), Retamal et al. (2011), and Willetts et al. (2010) fully applied this method. These studies determined the cost of the next additional unit of service provision for either households or users (see Criterion 5.1).

Criterion 6.2 evaluates whether a study adjusts prices to a common year and currency, explicitly citing conversion factors. As indicated in Table 2, this criterion fully applied to most of the included literature.

Criterion 7.1 assesses the effectiveness of economic models in representing the specific contexts of urban sanitation interventions within cities of LMICs. Models that fully apply this criterion need to scale and integrate data accurately to forecast cost and benefit changes over time. This will require reconciling fragmented data and ensuring assumptions align closely with real-world behaviours and their impact on demand for urban water resources.

Table 2 indicates that resource flow models most effectively apply Criterion 7.1, as they more closely determine the broader economic impacts of urban sanitation interventions. Six reviewed studies utilised water balance models to estimate economic costs in alignment with this intention. Retamal et al. (2011) and Willetts et al. (2010) demonstrated a rigorous approach to meeting this criterion by conducting a residential water end-use analysis model based on representative household surveys. This method enables the bottom-up prediction of wastewater flows over the entire timeframe of their analysis, derived from end-use practices within households, such as bathing or flushing. This approach allows the model to test the impact of context-specific assumptions on urban sanitation interventions from diverse perspectives (Retamal et al. 2011) (see Criterion 1.2).

Other studies either mostly or partially apply this criterion, depending on the rigour of the model applied. For instance, studies may represent resource flows using generalised per capita assumptions about total residential wastewater production (Jung et al. 2018; Manga et al. 2020), empirical cost functions (Peal & Georges 2013), or existing standard design parameters (Mburu et al. 2013; Shi et al. 2018) to predict the future costs of urban sanitation interventions. Studies that evaluate non-sewered urban sanitation options adopt resource flow models that estimate volumes of faecal sludge, as opposed to wastewater. For example, Tilmans et al. (2015) evaluated a container-based sanitation option using a model derived from empirical measurements. In contrast, Furlong et al. (2020), Kennedy-Walker et al. (2016), and von Münch & Mayumbelo (2007) determined resource flows based on per capita faecal sludge generation rate estimates derived from other contexts.

An alternative method of modelling the costs of urban sanitation interventions that partially applies Criterion 7.1 is financial flow modelling, as adopted by Carrard et al. (2021), Dodane et al. (2012), and McConville et al. (2019). These studies utilise historical financial data from the records of utilities, local authorities, and other service providers to project future costs. However, while the analysis facilitates a valuable financial comparison, the approach does not meaningfully account for the physical impacts of resource flows from changes to populations over time. This impacts how accurately they represent the actual context of the study.

Criterion 8.1 assesses whether studies clearly define and justify the applied discount rates and time horizons. Most studies fully applied this criterion, typically applying discount rates of 4–12% across time horizons of 10–30 years.

Criterion 9.1 evaluates whether confidence intervals are included for stochastic data used as model inputs in comparisons of urban sanitation interventions. Although this was not common in the reviewed literature, this criterion helps decision-makers to clearly understand the level of certainty of the results of an economic analysis. Confidence intervals indicate the variability parameters such as unit costs and resource quantities, which can vary widely across different urban settings and result in significant errors and poor decision outcomes.

An example that fully applies Criterion 9.1 is Jung et al. (2018), who defined confidence intervals for various input variables, including the discount rate, cost of electricity, and the degree to which modelled wastewater treatment configurations were decentralised. These intervals inform more certain decisions about the planning of wastewater treatment configurations in this context. Other examples partially apply Criterion 9.1 by providing confidence intervals only for critical variables with uncertain values, for example the sale price of reuse products (Carrard et al. 2021), the mass of faecal waste produced per day (Tilmans et al. 2015), the willingness to pay for sanitation services, Delaire et al. (2020), and an overall outcome measure of cost per death averted (Meddings et al. 2004).

Criterion 9.2 evaluates how sensitivity and scenario analyses enhance the use of modelled data to address uncertainties. Seven studies fully apply this criterion, as detailed in Supplementary Table S1, although they apply these tools in varied ways.

Sensitivity analysis tests assumptions about significant input parameters over a defined range. Some analyses focus on a key parameter, such as the discount rate or time horizon of a study (Prihandrijanti et al. 2008; Dodane et al. 2012; Willetts et al. 2010), the cost of fuel or electricity (Jung et al. 2018; Manga et al. 2020), the level of subsidy provided (Delaire et al. 2020), the sale of reuse products (Schuen & Parkinson 2009; Carrard et al. 2021), or non-monetary benefits, such as averted health costs (Hutton et al. 2014). Other analyses are more comprehensive and are used to determine the parameters with the most potential to impact a decision (Hutton et al. 2014; Jung et al. 2018; Ketema & Langergraber 2015).

Scenario analysis poses what-if questions to a model to consider the impact of assumptions on a model over time and how this scenario may impact multiple parameters at the same time. This method is primarily used to appraise the impact of uncertain future scenarios when comparing sanitation service options (Prihandrijanti et al. 2008; Kerstens et al. 2009; Schuen & Parkinson 2009; Kennedy-Walker et al. 2016; Jung et al. 2018; Furlong et al. 2020). Other applications are more specific, seeking to validate a cost function model (Ketema & Langergraber 2015), test and optimise assumptions about the population assumed to be serviced by a service provision model (Carrard et al. 2021; McConville et al. 2019), or structure lifecycle costs over time with greater certainty (Willetts et al. 2010).

Criterion 9.3 assesses the use of deliberative methods in urban sanitation studies to address uncertainties from diverse perspectives. This criterion highlights the importance of incorporating a broad range of viewpoints to enhance the equity of economic analyses and prevent biased decisions in service provision. Intentionally including these perspectives helps align studies with CWIS principles (Lüthi et al. 2020), fostering deeper reflection on and addressing the root causes of disparities in service access.

While no studies have fully applied Criterion 9.3, some enabling practices with the capacity to enhance the participation of traditionally excluded groups in economic appraisals were identified. For example, Willetts et al. (2013) implemented a sustainability assessment that brought together various stakeholders to discuss complex issues of social equity and environmental sustainability, prompting discussions about challenges in reaching consensus. Similarly, Dahiya et al. (2006) employed an informed choice approach, enabling informal households to actively evaluate trade-offs between costs and the intangible benefits of different service options. Future methodologies should further these practices, ensuring comparative economic analysis shapes inclusion in urban sanitation planning decisions.

Criterion 10.1 assesses whether an economic analysis measures outcomes in incremental units of TACH or TACC (see Criterion 5.1) and compares these against an existing least-cost without-project standard (see Criterion 2.1). Among the studies reviewed, only Kennedy-Walker et al. (2016) fully applied Criterion 10.1 by conducting an incremental analysis. This study compared faecal waste transportation and treatment costs to those of a least-cost pilot study in comparable units of total monthly cost per household.

Criterion 10.2 evaluates whether economic appraisals of urban sanitation report aggregated outcome measures that are then disaggregated to detail significant cost and subgroup perspectives (see Criterion 1.2), lifecycle cost components (see Criterion 6.1), and stages of a sanitation value chain. Studies that fully apply Criterion 10.2 include Carrard et al. (2021), Delaire et al. (2020), Dodane et al. (2012), Hutton et al. (2014), and McConville et al. (2019).

Criterion 10.3 extends the principles of Criterion 10.2 by evaluating whether economic analyses support flexible interpretation from diverse urban perspectives and scales. This criterion stresses the importance of transparent and non-prescriptive reporting, avoiding sole reliance on quantitative measures. It advocates for an iterative and collaborative analysis process that adapts to the diverse and evolving economic relationships among various cost and subgroup perspectives.

While current practices often fall short of involving participants directly in analysing results, several studies illustrate emerging approaches that enhance collaborative decision-making. For example, Delaire et al. (2020) suggested that considering the spatial heterogeneity of cities could refine economic analyses, optimising sanitation costs for different urban contexts. Schuen & Parkinson (2009) and Peal & Georges (2013) demonstrated how what-if scenarios may be used to explore how changes in urban sanitation configurations might impact different cost perspectives. In addition, Furlong et al. (2020) conducted participatory workshops to ensure that a financial flow model adapted from another context resonated with the diverse experiences of users and service providers before its implementation in economic analysis. Future methodologies should develop tools that allow decision-makers across an urban spectrum to relate and interpret results on their own terms. This may facilitate a more inclusive and comprehensive understanding of economic analyses, aligned with CWIS principles.

This review examined how methods from costing studies over the past two decades align with evaluative criteria for least-cost comparative economic analysis of CWIS interventions in LMICs. By analysing how each criterion was applied, this paper provides valuable insights for planners aiming to design comparative economic analyses to select equitable and sustainable urban sanitation configurations.

Several challenges emerged from applying these criteria. Local authorities will need practical support to reframe economic analysis from first principles to an inclusive societal cost perspective. This support should facilitate defining a problem that includes all relevant subgroup perspectives in an analysis. It will require negotiating a collaborative definition of an existing without-project standard for sanitation service provision and collecting context-specific data for meaningful comparison. CWIS principles imply the need to take calculated risks to transition from focusing on existing infrastructure options to focusing on a diversity of approaches and outcomes. This will require longer-term economic analysis methodologies to compare context-specific innovations that respond to persistent unmet service demands. Decision support models should aspire to engage service users in this process critically.

The primary contribution of this review is highlighting opportunities to enhance economic analysis methodologies to better address the diverse needs for sanitation in cities in LMICs. The evaluative criteria require broader discussion within the urban water and sanitation sector to improve how they are defined and weighted in an overall evaluation. This activity should be coupled with additional context-specific economic analysis studies aligned with these criteria, complementing more generic heuristic comparisons.

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

The authors declare there is no conflict.