Towards More Equitable Urban Greening: A Framework for Monitoring and Evaluating Co‐Governance

Urban greening has grown in significance in Europe and worldwide as a presumed “public good” initiative, delivering a range of benefits for human health and wellbeing. To redress inequalities in the distribution of such benefits, attention has turned to the potential of collaborative governance. Indicator‐based frameworks have also begun to receive attention for their ability to monitor and evaluate not only the performance of greening interventions


Introduction
In the face of growing threats to quality of life in cities worldwide-among them global warming, pandemics, and societal fragmentation-urban greening has captured the attention of urban planners and managers for its potential to deliver a range of benefits (Thompson, 2002).Green spaces have a key role to play in mitigating heat (Rahman et al., 2020) and regulating stormwater disposal (Zölch et al., 2017), while access to nature contributes to better mental and physical health (Tzoulas et al., 2007).However, as urban greening has gained momentum, attention has increasingly turned to the recipients of these benefits, and related questions of social and environmental justice (Anguelovski et al., 2020;Bauer, 2023;Connolly, 2019;Rutt & Gulsrud, 2016).
To make sense of the different dimensions of justice, many scholars have applied a tri-partite concept, composed of distribution, recognition, and procedure (Fraser, 1998;Schlosberg, 2007).Distribution concerns uneven impacts from environmental burdens and availability of environmental goods (Liotta et al., 2020).Similarly, the formal and informal procedures for making decisions related to urban greening are not readily accessible to everyone, such that decisions inevitably privilege certain interests, while neglecting others (Anguelovski et al., 2019;Baasch, 2020).Recognition justice supports the other dimensions in that neither equitable processes nor outcomes are possible without targeting differences in resources, capacities, and needs among individuals (Fraser, 1998;Young, 1990).More recently, Anguelovski et al. (2020) have pointed to the failure of the tri-dimensional framework to adequately reflect the lived experience of injustice.They propose an analytical approach based on principles of emancipation, anti-subordination, intersectionality, and (feminist) relationality, which they argue must fundamentally underpin urban greening in order to achieve justice.Their criticism appears to stem at least in part from a tendency for previous research to focus on one or the other dimension, commonly distribution, rather than recognise the three as interdependent (Bauer, 2023;Bennett et al., 2019) as well as failing to address power relations.We maintain that the dimensions of recognition, procedure, and distribution, when considered together, remain a useful lens through which to characterise an otherwise abstract concept, particularly in the context of practice-engaged research.Evidence suggests that all three dimensions are rarely reflected in the criteria that municipal decision-makers use to determine priorities for urban greening (Fisher et al., 2021;Hansen et al., 2022;Hoover et al., 2021).

Theoretical Background
Training a justice lens on urban greening projects demands that we move beyond technical requirements and recognise their fundamentally political nature, paying attention to the governance arrangements that underpin their planning, design, and stewardship (Hansen et al., 2022;Patterson et al., 2017;Pauleit et al., 2017).The concept of governance inherently recognises that formal government is not solely responsible for managing affairs, but engages with a range of institutions (Healey, 2006) and non-state actors (Emerson & Nabatchi, 2015).For our analysis, we take collaborative governance (Ansell & Gash, 2008;Emerson et al., 2011) as our framing concept, in response to the deficits in collaboration identified as key barriers to delivering urban greening (Frantzeskaki, 2019;Mekala & MacDonald, 2018).We define collaborative governance as a process of actors across the public, civil society, and private domains working together towards shared objectives for positively and equitably transforming the urban environment through the planning, design, implementation, and management of a nature-based intervention (Lim et al., 2023, after equity is important, since, while collaborative governance is sometimes assumed to translate to equitable outcomes, empirical evidence is limited (Dobbin & Lubell, 2021;Toxopeus et al., 2020).To move towards making this definition operational, we rely on five normative principles (see Table 1), defined through a systematic, qualitative literature review (Lim et al., 2023).
In Europe, a number of practice-based research projects have sought to guide collaboration on urban greening interventions, often framed as nature-based solutions (NbS), a concept that has benefited from substantial European Commission research funding (Bradley et al., 2022;Collier et al., 2023;DeLosRíos-White et al., 2020;Hölscher et al., 2024).As part of such studies, indicators have typically been deployed to assess the benefits of greening (Dumitru & Wendling, 2021).Just like quantifiable impacts, e.g., a reduction in temperature or air pollutants, indicators are capable of demonstrating a shift to a more collaborative way of governing, allowing targets to be set and progress tracked (Bennett & Satterfield, 2018;da Cruz & Marques, 2017), as a basis to redirect resources where most needed (Dumitru & Wendling, 2021;Morgan et al., 2022;van der Jagt et al., 2022).Indicators can also radically simplify complex information for diverse audiences: a powerful tool for policymaking (Keirstead & Leach, 2008) and demonstrating success Table 1.Normative principles underpinning successful collaborative governance of urban greening (Lim et al., 2023) and links with dimensions of justice.

Definition Justice dimensions
Collaborative Seeking out and mobilising governmental and non-governmental actors from multiple disciplines, departments, and levels; coordinating individual efforts to solve common problems adequately and at an agreeable cost (Börzel & Panke, 2007;Frantzeskaki & Rok, 2018).

Legitimate
Trust in and acceptance of decisions is ensured by adhering to democratic norms: i.e., equal participation of those affected; fair, transparent, and accessible decision-making processes; and accountability (Buizer & Van Herzele, 2012;Secco et al., 2011).
Participation and outcomes based on democratic norms (procedure; distribution)

Adaptive
Planning and implementation are strategic, open-ended, and iterative, involving continuous reflection, and learning from feedback loops to improve processes and outcomes (Martin et al., 2021;Mok et al., 2021;Morgan et al., 2022).
Efforts to continuously improve participation and outcomes (procedure; distribution)

Empowering
Equipping less powerful actors with the agency to assert their interests, based on a dynamic, evolving process of deliberation in which public, private, and civil society actors are afforded meaningful opportunities to create and share knowledge, challenge existing ideas, and proportionally influence outcomes (Barletti et al., 2020;Morgan et al., 2022).
Inequalities in resources recognised (recognition; distribution); targeted efforts to support participation (procedure)

Responsive
Actively recognising and analysing the specifics of local context at the outset of decision-making, with attention to differences in needs, interests, and values between and within communities; identifying pressing local challenges, understanding locally-specific institutional arrangements, and enabling local and indigenous knowledge to enter the process (Baasch, 2020;Graham, 2015).
Difference and plural, local knowledges recognised (recognition); participation actively enabled (procedure) (Astleithner et al., 2004;Emerson & Nabatchi, 2015).However, to deliver on this potential, the indicator set must be fit for its intended purpose.Da Cruz and Marques (2017) assert that indicator sets for evaluating municipal governance should be concise, as simple as possible, and clearly related to an objective.The set should also be exhaustive, while free from redundancy, discretionary indicators, or any factors that cannot be influenced by local government.In practice, meeting these conditions involves trade-offs, e.g., exhaustiveness will likely compromise concision and simplicity.Deciding which trade-offs are acceptable depends on the purpose.While many studies have looked at ways to create indicator sets that are optimal for the purpose of delivering robust scientific data, our study aligns rather with scholarship on their potential to build capacity and open up new discussions and modes of working (Bennett & Satterfield, 2018;Holman, 2009;Zafra-Calvo et al., 2017).
Evaluating collaborative governance faces several challenges, as highlighted by Emerson and Nabatchi (2015)  different agencies; difficulties identifying causal links in complex systems (especially when people and their behaviour are the subject); complex and detailed monitoring requirements; insufficient alignment with local policy agendas; and incompatibility with the realities of working environments (Dumitru & Wendling, 2021;van der Jagt et al., 2022).Uptake of indicators in decision-making may also depend on political will, established working practices, and the values and interpretations of individuals (Astleithner et al., 2004;Holman, 2009;Rydin et al., 2003).Aside from the question of uptake, which indicators and analytical scale are chosen can influence their potential to help transform governance practices (Astleithner et al., 2004;Beck et al., 2021;van der Jagt et al., 2022).Indicators that demonstrate progress only for a localised site or time-limited project are unlikely to drive change at a broader, city-wide scale; a problem often observed in urban greening projects seeking to stimulate public participation (e.g., Willems et al., 2020).

Method
This study took place as part of the practice-based research project JUSTNature, aiming to foster a just, low-carbon transition in European cities. Six city administrations and one city-owned company are involved, with their representatives working with researchers to design and implement nature-based interventions.
The cities vary in size, represent a range of biogeographical and socio-economic profiles, and face different challenges related to collaboration, offering highly diverse settings from which to examine transformative urban greening (see Figure 1).Our focus is not just the technical design and delivery of interventions, but also experimentation with different approaches to engaging stakeholders in their design, implementation, and stewardship.An indicator-based framework offers a means to monitor and reflect on these efforts, with a view to fostering a culture of collaborative governance in the longer term.
We built on a key review of existing assessment frameworks and guidance on participatory indicator selection (van der Jagt et al., 2022) to define a framework for evaluating governance of urban greening.
Our starting point was a compilation of existing key relevant indicator sets (Dumitru & Wendling, 2021), as well as additional indicators defined by van der Jagt and Buijs (2021).To these, we added indicators from empirical and theoretical studies (Kabisch et al., 2016;Morgan et al., 2022;Secco et al., 2011), resulting in a list of 51 indicators.We reviewed each indicator for relevance in the context of what could be measured through the project, discovering some shortcomings.Several indicators insufficiently described what was to be measured and how; very few indicators related to digital technology ( = 2), despite its growing role in shaping governance interactions; and a direct link with justice could be observed for less than half ( = 19).
For example, indicators for participation were often phrased in terms of generic citizens or community, without distinguishing between groups (absence of recognition justice).In light of these deficiencies, we saw scope to enrich this list by further characterising several indicators, and defining new indicators, derived from literature review (Lim et al., 2023).All indicators were then reviewed for their relationship with our five normative principles for good co-governance and assigned to one or more principles.In this way, we arrived at a set of 126 indicators capable of demonstrating progress towards co-governance according to the principles.
To translate our set into a reduced and meaningful number of indicators for each city, we drew on insights into participatory indicator selection from van der Jagt et al. ( 2022) who highlight the importance of seeking input from representatives of the contexts where monitoring should take place.We had laid groundwork through earlier engagement with the city partners during two online discussions-firstly introducing the concept of co-governance and secondly exploring its relevance to individual working environments-and a working session in person to discuss the principles in relation to local challenges (see Figure 1).We then reduced the list of indicators-an important step to lower the burden of participation, given that an overly lengthy list risked respondents disengaging from the content.Rather than maximise the number of indicators under consideration, our intention was to promote commitment within the lifespan of the project to working with a focused set of selected indicators, grounded in all five normative principles.Our primary exclusion criterion concerned indicators where data would already be collected through the project via established mechanisms (e.g., surveys developed to evaluate local workshops), and whose ranking was thereby redundant ( = 46), while ensuring that all principles continued to be represented.We then developed an online survey (see Figure 2

Ranking the indicators
To select indicators that are suitable to your context, we invite you to consider two ranking criteria: Relevance: Indicators are relevant if they are linked to the local objecঞves defined by your City Pracঞce Lab, to the interests of one or more stakeholder group, and/or to an exisঞng policy or legal framework, and if they are suitable for monitoring change over ঞme.
Feasibility: Indicators are feasible if there are adequate ঞme and resources available to carry out the assessment and monitoring process.This also means data of an adequate quality is available, and no specialist criteria is required to collect it.

Results
The distribution of rankings across cities varied greatly (see Figure 3).Of the 80 indicators, 76 were ranked very relevant by representatives of at least one city, irrespective of feasibility.Seventy were ranked very relevant and very feasible, while 62 were ranked very relevant, but only a little feasible or not at all feasible.
Considering potential transferability of the indicators between different local contexts, it is important to look at trends across the cities.No single indicator was ranked very relevant and/or very feasible by respondents from all seven cities, while only one indicator was ranked very relevant and very feasible by five cities.Seven indicators were seen as very relevant and very feasible by respondents from four cities, and 20 from at least three cities.This demonstrates significant variation in perceived relevance and feasibility across the cities, highlighting the importance of context.When only relevance is considered, the number of common indicators increases significantly.Three indicators were selected by six cities, 23 by five cities, 37 by four cities, and 57 by three cities.Respondents were asked to explain the reasoning behind their rankings.Their reflections provide useful insights into attitudes to relevance and feasibility (see Table 3 for an illustrative snapshot).While the reflections above provide insight into respondents' attitudes to working with the indicators, we found few clear explanations of why certain indicators were very relevant, although we had asked that each indicator be considered in relation to local objectives (Rows 3 and 7 hint at possible aims, but remain at a general level).In fact, the survey responses made clear that local objectives had typically either not been clearly defined, or were related to technical or environmental outcomes, such as enhanced biodiversity or cooling effect.We addressed this gap by reviewing internal project documentation gathered to date (stakeholder mapping, presentations, workshop reports, and discussions with city partners about perceived obstacles to collaboration on urban greening) to inductively define possible objectives linked to co-governance for each city, at both the scale of the intervention sites and more broadly at city level.
We then reviewed the indicators each city had ranked very relevant and feasible for coverage of our five  Not at all Very "More is not automatically better.We're talking about a public green space.E.g., LBGTQ+ organisations did not even answer the workshop invitation, as it is not about their issues.We must focus on who is affected with the project, not theoretical quotas that make no sense."(City 3) A little No answer "In a small town like ours, it is impossible to have stakeholders representing particular groups."(City 1) (3) Bottom-up initiatives-volume: Number of citizen initiatives, proposals.
Very Very "Active participation and initiatives from the citizens would be very interesting and beneficial.We would be building towards a better relationship with the community and to their understanding of governmental policies.

Very
Not at all "We can only measure this in the project, meaning we will not get a realistic picture of the situation in general.Such workshops only happen in international projects, where they are mandatory.We would be interested to have guidelines for how to hold workshops efficiently and justly in the future-because there is little experience, but an openness to doing more."(City 5) Very A little "The council doesn't have different departments but these exist at a national level.So far, we had several stakeholders representing the environmental protection segment but need more representation from different departments.It is very relevant to have these stakeholders on board to represent these different interests."(City 1) normative principles.In most cases, further indicators were excluded after defining an objective, while the "principles check" entailed some additions (e.g., for Gzira, an indicator related to the principle "adaptive" was missing; three selected indicators could not be directly linked to an objective and were excluded; while two new indicators were suggested in their place).Proposed objectives and corresponding indicators were presented to each city partner for review and adjustment (see example in Figure 4).
Limited municipal culture of collaboraঞon.

Relaঞvely low level of ciঞzen engagement
Typically mostly white, older educated people parঞcipate-other groups are not engaged.Aiming for simplicity and concision (da Cruz & Marques, 2017), and mindful of resource constraints (Dumitru & Wendling, 2021), we further reduced the number of indicators by merging those that were deemed too similar to one another, taking into account feedback from respondents.This resulted in a total number of 48 indicators (see Supplementary File), with each city-specific set between 8 and 28.The wide variation is explained by the varied results of the ranking exercise.Our final city-selected indicator set ( = 48) was made up of 16 from existing sets of indicators (Dumitru & Wendling, 2021;Kabisch et al., 2016;Morgan et al., 2022;Secco et al., 2011;van der Jagt & Buijs, 2021), and 32 derived from literature review.The relatively high proportion derived from literature review suggests that expanding on existing indicator sets was useful to improve perceived relevance and feasibility, while the presence of a slight majority of indicators capable of monitoring dimensions of justice ( = 26, or 54%) is a significant proportional increase on the 51 indicators we started with ( = 19, or 37%).Figure 5 outlines the process and a quantitative overview of results.

Perceived Relevance and Feasibility
We set out to define city-specific indicator sets seen as relevant and feasible by the people responsible for working with them.A majority of indicators were perceived as relevant in at least one city, with only four out of 80 indicators not ranked very relevant at least once.However, when results are averaged across all cities, only about half the indicators were viewed as relevant, and not a single indicator seen as very relevant by respondents from all seven cities.Although indicators are necessarily locally specific, such a low level of commonality is nonetheless remarkable.Feasibility also came to the fore as a key barrier to using the indicators, with on average only 25% of all indicators seen as both relevant and feasible.In studies where comparability between cities is critical, a productive avenue may be to explore the specific reasons behind difficulties collecting data and work out strategies to overcome these.
Respondents questioned how certain indicators could be measured, meaning feasibility may have been ranked lower where there was doubt about what data to collect or how.They also raised concerns about the workload and skills associated with collecting data, suggesting that feasibility may also have been ranked low as an expression of a respondent's concern that this burden would fall to them.Crucially, several indicators were flagged as not within the respondent's capacity to influence.This suggests that relevance and feasibility may not sufficiently characterise the local validity of an indicator, where agency (real or perceived) of the individuals responsible for monitoring is lacking.While van der Jagt et al. ( 2023) have identified agency as a key category for monitoring co-governance of urban greening, our findings rather relate to "meta-level" agency, i.e., less so an indicator that measures availability of institutional support and resources (although this is important), but rather that the individuals responsible perceive that they can influence all indicators.

Challenges in Contextualising and Politicising an Indicator-Based Framework
Having concluded a key phase in setting up a context-specific framework for monitoring and evaluating co-governance, we can identify some early achievements.A future challenge is to politicise this framework, i.e., to ensure that the monitoring process integrates entry points for broader change beyond the project scale (Turnhout et al., 2020)-a frequent shortcoming of practice-based research projects engaged in experimentation (von Wirth et al., 2019;Voytenko et al., 2016).
It is likely that this will play out less at the level of the indicators themselves, but rather through their corresponding objectives.In this regard, we have advocated for both site-level and city-level objectives, and provided guidance to the city partners on how to engage colleagues in structured discussions about their own objectives, with a view to anchoring the project activities in a shared strategic context with longevity.
Our results so far suggest that underscoring the agency of individuals within the project will be critical so as to ground ambition in practical, achievable actions, without losing sight of potential for broader change.
Further analysis could build on studies that have explored the specific capacities needed to foster longer-term shifts to collaborative governance, and the extent to which these manifest over the course of the project (Hölscher et al., 2024;Wolfram et al., 2019).managerial roles rather than political ones, and in some cases engaged as contractors external to the city administration.Involving a wider range of respondents in future monitoring (Hansen et al., 2022) may well enable richer insights, and indeed be crucial to their explanatory power, since individuals within the same group may disagree on whether a collaboration was indeed successful (Emerson & Nabatchi, 2015).

Limitations and Future Research Directions
The real test of our indicator framework lies in its future implementation.We are already alert to concerns raised by city partners about the workload of data collection, and recognise that perceived feasibility to collect data for one indicator cannot necessarily be extrapolated to a whole set.Our work ahead includes developing guidance on data collection through surveys and interviews, as well as a mechanism to periodically self-reflect on progress towards objectives.Our experience to date suggests that this needs to be accompanied by support from the local research institutions in each city, along with continued efforts to demonstrate the value of monitoring and evaluating co-governance and to illustrate the agency of city partners.

Conclusion
Our experience working with city partners to define a framework for monitoring and evaluating co-governance of urban greening has generated some key observations that will guide our further work and may be useful to other research teams.A key achievement has been to develop an expanded set of 126 indicators, structured according to five normative principles for collaborative governance, and to test 80 of these for their relevance and feasibility in local contexts.We found that organising indicators according to normative principles and applying the lenses of relevance and feasibility was helpful to make a complex concept operational.While the indicator ranking exercise was primarily intended to contextualise the indicators and define city-specific indicator sets for further use, the results were pivotal in providing a basis to concretely define objectives for local collaborative governance.These findings add support to the case by others for more participatory approaches to creating assessment frameworks for urban greening (van der Jagt et al., 2023).
The ranking exercise also brought to light some reasons that existing indicators for monitoring co-governance of urban greening might not support transformative change in practice.Some barriers can be observed through our ranking criteria of relevance and feasibility, e.g., difficulty interpreting a generalised indicator as relevant to local conditions, and expected difficulties collecting data.We also found evidence of a varied sense of agency to influence the measures of progress among those responsible for monitoring it-particularly concerning potential change beyond the project boundaries.While relevance and feasibility are sound preconditions for a contextualised indicator set, if engagement with the political landscape beyond the project is sought, then capacity-building efforts are also needed to communicate the value of monitoring collaborative governance, and foster awareness of individual agency among those responsible.

Figure 1 .
Figure 1.Overview of the cities, their representatives, urban greening projects, and key challenges for collaborative governance identified to date.
A[er selecঞng the ranking level (3 = very, 2 = a lile, or 1 = not at all), please explain your choice, referring to your local context, challenges and objecঞves.Indicator: Parঞcipaঞon: Degree of Parঞcipaঞon Descripঞon: Following Arnstein's ladder of parঞcipaঞon (ranging from manipulaঞon to ciঞzen control), parঞcipatory processes (e.g.workshops) are planned and evaluated by both parঞcipants and municipality How can you measure it?Survey, Interview, Focus Group Relevance: How relevant is this indicator to your local objecঞves?Please explain why Feasibility: How easy is it to collect this data?Please explain why

Figure 3 .
Figure 3. Results of the ranking exercise.

Figure 4 .
Figure 4. Sample of two local objectives (as proposed by authors) and corresponding indicators for one city.

Figure 5 .
Figure 5. Overview of developing city-specific indicator sets.
First, we have brought together indicators on co-governance of urban greening from dispersed sources, and structured these according to five principles, ensuring that individual indicator sets reflect a range of normative underpinnings of successful co-governance.Second, the process of ranking indicators for relevance revealed a cognitive gap between the research team's ambitions for steering a governance shift and the perspectives of city partners, demonstrated most significantly by difficulties linking indicators with clear local objectives.Exposing this gap has made it possible to bring a more structured and strategic basis to future monitoring, by defining local objectives for collaborative at site and city level and thereby enabling scrutiny of progress; much like environmental or technical benchmarks would typically be evaluated(Bennett & Satterfield, 2018;da Cruz & Marques, 2017).Related discussions and other engagement activities with the city partners on indicators before and since the ranking exercise also provided key lessons for our research design.For example, comments about lack of clarity and perceptions of similarities between indicators prompted us to reformulate some indicator descriptions, as well as to elaborate guiding questions and stepwise checklists clarifying how data can be collected.Our experience also points to challenges engaging participants in contextualising indicators.First, although we align with scholars who have identified a need to expand existing indicator sets related to governance of urban greening (van der Jagt et al., 2023), we found a large number of indicators makes participatory selection difficult.Our reduced list of 80 was still perceived as overwhelming, and subtle differences between certain indicators dismissed as trivial.We support the view of van der Jagt et al. (2023) that comprehensiveness must be balanced with concision.Second, the generalised wording of some indicators offered substantial room for interpretation.As noted byvan Oudenhoven et al. (2018), a problem with seeking to develop transferable indicators is that indicators often need to reflect very specific policy issues to guarantee relevance.Some indicators may have been rejected where relevance needed to be made clearer, e.g., through examples of what the indicator could mean in practice.The impact on our results we consider insignificant, given that we aimed to promote commitment to a focused set of selected indicators rather than maximise the number of indicators in use.However, there is scope for future studies to improve accessibility of existing lists, e.g., clustering related indicators, providing (concise) guidance on how to adapt a generalised indicator description to a local situation, and further distinguishing between indicators with higher and lower demands for data collection.

A
limitation of study is that we have not yet determined exactly how suitable our indicator sets are to evaluate equitable co-governance of the greening projects.Such an analysis could deploy certain key requirements of a sound indicator set (feasibility, relevance, avoidance of redundancy, coverage of principles) to analyse and score each city's set of indicators, as well as checking that dimensions of justice are sufficiently represented.This latter aspect lies at present primarily latent in those indicators aligned with normative principles with links to dimensions of justice.Another possibility would be to weight individual indicators for their explanatory potential in connection with each principle, as a means of further reducing the indicator sets to only the most relevant, and in turn reducing the evaluation workload.Such analysis would benefit significantly from our city partners' input, based on their initial experience working with the indicator sets-not yet available at the time of writing.Our findings are not without bias, since we have worked predominantly with only one or two individuals from each city.The survey responses, documentation from which we defined local objectives, and resulting city-specific indicator sets represent a small group and how they perceive their local realities, and in turn how we have interpreted them.Perceived relevance, feasibility, and agency with respect to indicators depend on several factors, including the degree of support available from local research partners; disciplinary background; role, department, and length of time in each; seniority; and resources available to influence strategic decision-making.Encountering capacity limits is a typical situation in practice-based research, where often only one or two people are resourced by the project, typically in technical or Table 2 provides a snapshot of commonalities, outlining the 37 indicators ranked very relevant by respondents from at least four cities.

Table 2 .
Indicators ranked very relevant (disregarding feasibility) by respondents from 4+ cities, and corresponding principles.Dark grey highlights indicate a subset of indicators also perceived as very feasible by 4 cities.

Table 3 .
Selected reflections on the ranked indicators from respondents in five cities.

Table 3 .
(Cont.)Selected reflections on the ranked indicators from respondents in five cities.