University of Minnesota, USA
Georgia Institute of Technology, USA
National Center for Atmospheric Research, USA
Yale University, USA
Yale University, USA
(Authors listed in alphabetical order)
Negotiations are underway to set objectives and targets and establish a framework for implementing Sustainable Development Goals (SDGs), to be adopted by the United Nations in September 2015. SDGs differ from their precursors, the Millennium Development Goals, in that they are meant to apply universally to all countries. We define universality as the ‘appropriateness’ of goals, targets, and indicators for global adoption. Universality is particularly important for cities, as acknowledged in Urban SDG 11, which calls to “Make cities and human settlements inclusive, safe, resilient and sustainable.”
To determine universal indicators is no simple task. Tensions often flare in negotiations, as some nations point to their common but differentiated responsibilities while a lack of consensus prevails on definitions and performance metrics for urban sustainability (Hiremath et al., 2013, Lynch et al., 2013, Shen et al., 2011). Efforts to devise a core set of indicators, including the Bangalore Outcome and Sustainable Development Solutions Network (SDSN), have examined a small set of ‘universal’ indicators but have yet to address how universality applies to specific urban areas. The United Nations Statistical Commission recently released a review and ranking of the feasibility, suitability, and relevance of proposed SDG indicators. The UN-Stats analysis, however, reflects the perspectives of national statistics offices, which are often ill-suited to see the needs and understand the scale of the city.
Our research analyzes SDG targets, with a focus on developing Urban SDG applications in two very different cities: Atlanta, United States and Delhi, India. We selected these case studies to test proposed targets and indicators in three areas: transportation, participation and planning, and per capita environmental impacts. Using international best practice criteria for indicator selection (de Sherbinin et al., 2013), the case studies survey available data and explore the limits and potential to develop appropriate and comparable indicators for the Urban SDG.
We start by asking: 1) What data are appropriate for SDG indicators and who collects and manages this data? 2) How well do potential indicators capture the intent of the goal target? 3) What are the challenges and opportunities of universal Urban SDG indicators? Table 1 below applies these preliminary findings to relevant indicators recommended by SDSN.
Table 1: Preliminary recommendations on the Urban SDG indicators proposed by the Sustainable Development Solutions Network (May 2015) (Sustainable Development Solutions Network 2015) using initial findings from Atlanta and Delhi case studies
|Goal.Target||SDSN Proposed Indicators (2015)||Comparable data availability & universal application||Recommendations for indicators, data sources, or further analysis|
|11.2||25. Road traffic deaths per 100,000 population||Data on road fatalities is readily available although not always at the urban scale.||Annual traffic deaths can also be normalized based on VMT or residents plus jobs to more easily compare different types of cities and urban areas (Chavez and Ramasami 2011). The raw number of traffic deaths is sometimes used, especially in cities implementing Vision Zero initiatives.|
|11.2||58. Access to all-weather road (% access within [x] km distance to road)||Potential to retrieve this data globally through satellite or mapping sources. Determining a standard for classifying roads may be a challenge.||All-weather road access may have different implications in more developed cities producing complications in comparisons. Also, the mode of travel and the distance to desirable destinations is necessary to understand access fully. A related indicator to increase availability / utilization of mobile weather and road condition data globally could be considered to improve safety / mobility (Drobot et al., 2014).|
|11.2/11.6||67. Percentage of people within 0.5km of public transit running at least every 20 minutes.||Data on transit locations and service estimates will be dependent on the operating authority.||Performance target also fails to account for destinations accessible by transit. Additionally, considering the 0.5 km distance along the transportation network to transit stops will provide a truer understanding of access. Occupancy in addition to frequency important if to more fully address life cycle energy; safety, sustainability, and resilience goals [Gupta 2010, Chester and Horvath 2009].|
|11.3||68. [Ratio of land consumption rate to population growth rate, at comparable scale] – to be developed||GHSL for land consumption. No universal disaggregated bottom-up population dataset (GPW and WorldPop show best available)||No ideal universal target for land use efficiency. Desirable trajectory depends on initial conditions, which vary. Land use efficiency is not directly related to other types of resource consumption, which could be addressed by incorporating additional remote sensing datasets (building heights, nightlights, street density).|
|11.3||95. Domestic revenues allocated to sustainable development as percent of GNI, by sector||Not urban specific or disaggregated in a way relevant to the goal or target||Revenues will vary depending on development stage, and size of revenue may not determine success of outcome.|
|11.6||47. Percentage of wastewater flows treated to national standards [and reused] – to be developed||Global data set available but aggregated at country, not city, scale. Need standard definitions for treatment levels and performance targets (could be set at progress towards 100%).||Expand indicator to be more applicable in developed cities, perhaps addressing overall watershed impact|
|11.6||69. Mean urban air pollution of particulate matter (PM10 and PM2.5)||Global satellite data available for PM2.5.||Set performance targets by the World Health Organization standards.|
|11.6||71. Percentage of urban solid waste regularly collected and well managed||No global data set available. Need standard definitions for management and performance targets (could be set at progress towards 100%).||Measurement methods needed to avoid outsourcing environmental impact and waste treatment beyond the city limits; expand indicator’s applicability in developed cities|
This preliminary research is designed and implemented by an interdisciplinary team with backgrounds in engineering, environmental sciences, geography, planning, and policy. The cases presented are from cities planning for sustainability and climate change; they are not meant to be directly comparable or generalizable, rather they illustrate how the SDGs could be effectively applied.
Criteria to analyze data for indicators:
- Data requirements identifies information necessary to calculate and monitor the indicator
- Data availability assesses whether data is available and, if so, who collects it
- Current status and baseline provides available data and aspirational data needs
- Relevance and sensitivity of indicator refers to how well the indicator directly measures the target and indicator sensitivity to local contexts and global problems.
- Part of the goal addressed refers to which of the four components in the Urban SDG “inclusive, safe, resilient and sustainable” the indicator addresses.
Transportation (Goal 11, Target 2)
“By 2030, provide access to safe, affordable, accessible and sustainable transport systems for all, improving road safety, notably by expanding public transport, with special attention to the needs of those in vulnerable situations, women, children, persons with disabilities and older persons.”
Transportation systems are often described with multiple performance measures (e.g., road fatalities, asset preservation, vehicle miles traveled, and transit ridership) and an extensive set of indicators is readily available (for most cities) to evaluate transportation performance. There are two key difficulties for selecting an indicator for Target 11.2. First, countless existing transportation performance measures and second, the wide-sweeping target leaves space for interpretation. We isolate appropriate indicators based on the target’s language. For example, we use the words “safe,” “affordable,” “accessible” and “sustainable” to determine relevant indicators that can speak to the goal’s intent. The target’s explicit reference to expanded transit services and attention to at-risk populations underscore concerns for equity that should be tracked and assessed across the groups listed in the objective.
Three of the 100 indicators proposed by SDSN address transportation explicitly (see Table 1). We suggest refining and complementing SDSN’s effort with four indicators, shown in Table 2. These four are frequently used as transportation performance measures (in the United States, Canada, Japan, and New Zealand), yet there are few metrics that track benefits across populations, a task which would require significant resources. Transportation performance measures are commonly used, though data availability remains a challenge, especially in countries and cities with few resources. Data to measure the suggested indicators is available for Atlanta, but comparable data on vehicle travel and transit access is not readily accessible for Delhi, India. Transportation agencies are, in general, the stewards of data relevant to our suggested indicators because they are responsible for the performance of their respective networks. All transportation agencies, however, do not track this information, though there are global datasets that may be used to gauge performance towards these targets.
Table 2: Review of data and indicators for transportation target in Atlanta and Delhi
|Indicator||Data requirements||Data collection and management||Current status/ Baseline||Value/ Relevance/ Sensitivity of indicator|
|Safety||Total transportation- related fatalities (per million-persons or per X VMT)||Metropolitan region data available, Atlanta Regional Commission||Union territory data, National Crime Records Bureau||452 in 2012 (Atlanta Regional Commission 2014)||1,831 in 2013 (National Crime Records Bureau 2013)||Baseline of transportation safety.|
|Accessible||Change in mode share: % of trips by public transportation||Metropolitan statistical area data for work trips available, American Community Survey||Union territory data for all trips available, Government of National Capital Territory of Delhi||3% in 2013 (US Census Bureau 2013)||31% in 2008 (Land Transport Authority Academy 2014)||Describes change in transit use and suggests changes in accessibility.|
|Sustainable||Average daily VMT or VKT per capita||Metropolitan region data available, Atlanta Regional Commission||Traffic data may not be available for Delhi.||28 (average daily) in 2012 (Atlanta Regional Commission 2014)||8.8 VKT/resident/day (Central Road Research institute 2009)(VKT/day almost doubled from 2002-2009)||Estimates the level of driving in personal vehicles and can be used for estimating other measures such as emissions and fuel consumption.|
|Affordable/For All||% of income or expenditures to transportation||Housing and Transportation Affordability Index, Center for Neighborhood Technology||Household Consumer Expenditure Survey in Delhi||54% for housing + transport; 23% for transport alone||5.4% housing; 9.4% transport (Food: 38%) (Gov’t of NCT of Delhi 2007)||Estimates average % of expenditures on transportation, housing and food (for case of Delhi) and relative to average household income (for Atlanta) to assess affordability and equity.|
Planning and Participation (Goal 11, Target 3)
“By 2030 enhance inclusive and sustainable urbanization and capacity for participatory, integrated and sustainable human settlement planning and management in all countries.”
The target calls out both the urbanization process—the mechanisms and institutions of land management, urban financing, and planning—and built products—the urban form, structure, and infrastructure systems that lock in the resource intensity of the built environment and constrain sustainability outcomes.
Inclusive and participatory are two of Target 11.3’s keywords that address planning processes; they are concepts that call for careful evaluation, as they are embedded in local history and culture. Strength indicators (high, medium, low) are one way to gauge procedural rights, including access to information, participation, and justice. However, Meaningful participation, does not ensure that decisions benefit all facets of sustainable development, often contained within the 3 E’s of environment, equity, and economy. Both Delhi and Atlanta have strong laws to protect freedom of information and participation. Implementation in urban planning processes, however, is often weak and access to decision-making is limited, especially for disadvantaged populations most affected by public policies (Tompkins and Adger 2004). A recent study on equity found Atlanta to be one of the worst cities in the U.S. for young people to better their circumstances and rise out of poverty. Indicators of inclusivity and participatory planning are critical to address equity for sustainable cities. Measuring input to planning processes is somewhat easier than measuring the long-term output of those decisions—integrating the two in the Urban SDG’s implementation will be a data and methodological objective.
Target 11.3 aims to facilitate the design and construction of sustainable built environments, stronger institutions, and protected natural systems, while accommodating increasing numbers of urban residents. Monitoring enhancements in the urban built environment come with the added benefits of mitigating global environmental problems, particularly climate change and loss of biodiversity. Urban areas account for 67-76% of global energy consumption and 71-76% of global greenhouse gas emissions, making them key players in climate mitigation efforts. The International Panel on Climate Change describes the key physical characteristics [IPCC 2014] of low carbon development patterns, identified as candidate indicators for SDG target 11.3: (1) high population and employment densities that are co-located; (2) compact urban form; (3) mixed land uses; (4) high connectivity street patterns; and (5) destination accessibility to jobs and services.
SDSN proposes an indicator under Target 3 on the ratio of land consumption rate to population growth rate, at comparable scale. There is no ideal universal goal for land use efficiency, and desirable performance trajectories depend on initial conditions, which vary according to location. Satellite imagery comprises universally available data describing land-use, but remote sensing science has historically focused on delineating the boundary between urban and rural land cover—not identifying changes in the built environment within urban areas. Indicators that solely describe urban land expansion may have easily accessible data, yet these figures are not directly related to low carbon development patterns. When available, high resolution imagery is more valuable for classifying inter-urban characteristics
Novel open-source data products, including the Global Human Settlement Layer (the extraction of built area density at fine resolution within urban areas), the VIIRS Day/Night band daily product (for capturing neighborhood-level energy demand patterns), SeaWinds Scatterometer data and ASTER GDEM (for vertical urban measurements), and OpenStreetMap (for interurban street layout, road density, and connectivity) provide new opportunities to develop indicators that directly relate to the Urban SDG targets.
Environmental Impacts per Capita (Goal 11, Target 6)
“By 2030, reduce the adverse per capita environmental impact of cities, including by paying special attention to air quality and municipal and other waste management.”
Data requirements and indicators needed to form a comprehensive picture of city environmental impacts vary widely (see Table 1). The diversity of data will require participation from many city actors, including from multiple infrastructure sectors (e.g., energy, water, buildings, transportation, etc.). Coordinating disaggregated data from disparate sources and management parties will likely pose a major challenge for many cities. Indicators that measure sectors with a strong informal aspect, such as solid waste management and recycling, will prove especially difficult to attain, particularly for cities in developing countries. These challenges beg the question: will the final Urban SDG targets measure and monitor only what is convenient? Or will indicators be aligned with local city priorities in order to increase the value of and motivation for data collection in each city?
Endeavoring to meet indicator objectives, cities may outsource their environmental burdens beyond municipal boundaries. Beijing, for example, recently announced that it would close all coal-fired power plants, a move widely applauded as an important step towards curbing carbon emissions and improving the city’s notoriously poor air quality. The effort’s true value, however, is uncertain considering that Beijing will compensate for the in-city production shortfall by importing energy from coal plants in outside provinces, shifting the environmental burden to a different location. Similar problems hold true for solid waste management. In Delhi the majority of the 70% of municipal waste that is considered “regularly collected” is dumped in areas on the city outskirts, without any treatment or environmental protection [Talyan et al., 2008]. Standardizing environmental impact accounting methods for all cities would mitigate urban pollution outsourcing. Trans-boundary accounting, for example, ensures that all GHG emissions and environmental impact are attributed to the location of the responsible user, regardless of where the emissions or impacts occur. Table 3 presents generally accepted environmental indicators (supplementing Table 1) applied to Delhi and Atlanta, to further develop understanding of data needs and differences in application from city to city.
Table 3: Review of data and indicators for per capital environmental impact targets in Atlanta and Delhi
|Indicator[Urban SDG Campaign 2015]||Data requirements||Data collection and management||Current status||Value of indicator||Concerns/suggestions/commence|
|Percentage of solid waste collected and managed||Ratio of waste generated to collected and managed||Office of Solid Waste Service, City of Atlanta||Municipal Corporation of Delhi; (Talyan et al., 2008); Private companies||Comprehensive system (Office of Solid Waste Service)||70-80% collected with questions of management; (Talyan et al., 2008)||Waste management||Concerns of outsourcing environmental impact|
|Concentration of fine particulate matter||PM2.5||AirNow; Georgia Department of Natural Resources||National Air Quality Index||Frequently above WHO standard of 10 μg/m3 (GDNP)||Range 54.3 to 338.7 μg m3(Tiwari et al. 2012)||Human and environmental health||Applicable in developing and developed cities|
|Percentage of waste water treated||Ratio of water supplied to water treated||Office of Water Treatment and Reclamation||Delhi Jal Board||Comprehensive system (OWTR)||61% treatment capacity (Kuar et al.)||Human and environmental health||Expand applicability to developed cities, perhaps addressing overall watershed impact|
|GHG emissions, tons/capita||Diverse data requirements across sectors||Data housed across departments; Office of Sustainability, City of Atlanta||Data housed across departments||2.682 t CO2 e/capita (Brookings Institute 2005)||2.3 t CO2e/capita (Chavez et al., 2012)||Value dependent on accounting methods||Coordination of methods needed; ; methods data intensive across sectors|
|Proportion of municipal waste recycled||Ratio waste generated to recycled||City of Atlanta Recycling Program||N/A||12.5% (as of 2012) (Recycling Program)||No data found||Resource management||Concerns of outsourcing environmental impact; difficult to collect informal sector data|
The Urban SDG provides an opportunity to track and compare indicators in different countries, helping urban leaders redefine local sustainable development and tackle global scale environmental problems. Our preliminary findings from testing indicators in two case studies suggest that a comparative framework of core indicators and data is a prerequisite for the targets and performance-based goals to be universally adopted. Without appropriate indicators the UN runs the risk of adopting metrics with limited saliency, legitimacy, or credibility, which would inhibit implementation and adoption of the Urban SDG. Below are insights responding to our initial questions on data, intent, and universality of the Urban SDG:
1) What data are appropriate for Urban SDG indicators and who collects and manages this data? Data custodians wanted. The Urban SDG’s dual purpose to address local and global sustainability challenges requires different data sets depending on the problem. Most international bodies that provide global data collect national level data, which is not relevant at the spatial or temporal scale necessary for the Urban SDG. Few have the capacity to capture extensive city-level data and monitor annually as the SDGs intend. Even where data exists it is often scattered across institutions, departments, and agencies, complicating coordination, long-term tracking, and comparison (e.g., city greenhouse gas accounting methods and energy use benchmarks). Figuring out how to accurately measure transboundary environmental impacts remains a critical challenge, as highly porous urban boundaries encumber data collection. Instituting sound governance and building competencies to properly collect, manage, and analyze data require implementers to confront issues of ‘nested’ authority (Ostrom et al., 1999, Ostrom 1990), capacity (Elmqvist et al., 2013, Bai et al., 2010), finance, and appropriateness.
2) How well do potential indicators capture the intent of the goal target? Square pegs, round holes. Urban SDG targets are broadly written, leaving indicators up to debate depending on what part of the target is prioritized. The proposed one indicator per targets does not take into consideration the complexity of urbanization and multidimensional nature of the Urban SDG. Indicators for sustainable urbanization must monitor changes in the built environment and land cover, within the city core and at the periphery. Distinct methods and timetables for measurement exist – some of which extend beyond the 2030 deadline (i.e., inclusivity and equity dimensions). It would be a mistake to merely refashion existing tools, indicators, and accounting methods developed and applied in European and North American cities. Failing to include input from officials and urban actors across Asia, Africa, and Latin America, where smaller rapidly industrializing cities will dominate future urbanization (UNDESA 2014), could result in lack of participation and accountability, limited progress on goals, or even direct conflict as a result.
3) What are the challenges and opportunities of universal Urban SDG indicators? A universe of possibility. Universal data across cities gathered through surveys and statistical collection systems are sparse and vary in accuracy. Urban planning indicators will therefore need to rely on remotely-sensed data sources. A recent flood of new data applicable to sustainable urban planning has created opportunities to derive indicators that are sensitive and relevant to planning targets. Still needed, however, are innovative remote sensing-derived measures that proxy sustainable development indicators and move beyond binary land cover classification.
To move a strategic planning process forward for the Urban SDG, we need to ask:
- What are the possibilities for a core set of universal Urban SDG indicators and performance targets?
- What are the processes and who will participate in determining appropriate indicators at the city-level?
- What are the potential global and local impacts of setting performance targets, then collecting and monitoring the data across all cities?
Testing indicators in Atlanta and Delhi demonstrates that current data variability and lack of comparable methods hinders implementation of the Urban SDG on a global scale. Target vagueness and data constraints leave performance indicators open for interpretation exacerbating existing tensions between universal and appropriate local implementation. Ultimately, the SDGs are a political process and the current push to select 100 indicators to measure and monitor sustainable development arbitrarily confines the integration of science. It is increasingly important to critically analyzes of the SDG process and indicator selection is therefore increasingly important to assess implementation of the Urban SDG, avoid negative consequences, and determine what additional work needs to be done.
Acknowledgements: This research would not have been possible without support from the National Science Foundation Research Coordination Network on Sustainable Cities, input from Kristin Oloffson, and editing by Carlin Rosengarten.
(from left to right)
Dana Boyer is a PhD student of Science and Technology Policy in the Sustainable Cities group at the Humphrey School of Public Affairs, University of Minnesota.
Stefanie Brodie is a PhD candidate in Civil Engineering in the Transportation Systems program at Georgia Institute of Technology.
Joshua Sperling is a NSF PIRE Postdoctoral Fellow at the NCAR Urban Futures initiative in the Climate Science and Applications Program.
Header Image: Delhi, India (left) – Atlanta, Georgia (right). Credit: Joshua Sperling and Stefanie Brodie