Developing a green infrastructure equity index to promote equity planning

Highlights

• We develop a Green Infrastructure Equity Index to identify “equity voids.”

• The index can be used to identify and empower communities with high need.

• The index framework is a flexible planning tool designed to empower communities.

Abstract

The Philadelphia Water Department has committed to taking a green infrastructure (GI) approach to reduce stormwater runoff and prevent combined sewer overflow events. Promoting GI as a stormwater management technique in a city necessitates development of a more distributed urban environmental management system, through which the city's water department needs to coordinate with a wide range of public and private stakeholders, shifting power from the utility to these other stakeholders. We argue that distributed urban environmental management can lead to more inclusive outcomes but only if there is an intentionality about how funds are distributed, which communities are prioritized, how partners are chosen and cultivated, and which types of projects are implemented in which neighborhoods. We suggest the development of an equity index to help identify communities that would most benefit from GI investment as critical for equitable GI planning. Using Philadelphia as a test case, we develop a Green Infrastructure Equity Index, designed with the indirect benefits of green infrastructure in mind, to determine which communities could benefit the most from investment in GI based on their “equity void ranking”. We argue that developing a GI Equity Index provides a much more nuanced analysis of communities that takes into account the built environment as well as the underlying social and economic conditions. The GI Equity Index also allows for a shift in the way we define equity. In doing so, it (1) changes the conversation about equity in GI planning using careful data analysis that takes into account both socio-economic and built environment variables; (2) provides a visual tool that communities can use to understand underlying conditions and the existing placement of GI; and (3) serves as a framework that can be tailored to allow communities to weight their priorities, putting more power in their hands.

Introduction

Concerns about reducing stormwater runoff are tied to the widespread use of combined sewer systems in older American cities, in which even small rain events may overwhelm system capacity and lead to the discharge of raw sewage into waterways. The US Environmental Protection Agency (EPA) is pushing for cities to reduce the frequency of combined sewage overflow (CSO) events so as to comply with Clean Water Act regulations. The green infrastructure (GI) approach is gaining in popularity, and now seventeen cities, including Philadelphia, have explicitly incorporated GI into their agreements with the EPA for reducing CSO events. (For the complete list, see http://water.epa.gov/infrastructure/greeninfrastructure/enforcement.cfm).

The Philadelphia Water Department (PWD) has committed to taking a green infrastructure approach to reduce stormwater runoff as part of its plan to prevent CSO events. This means using a variety of natural and engineered vegetated practices to manage stormwater as opposed to building traditional “gray” infrastructure. Though the term green infrastructure is often used broadly to refer to vegetated land uses including parks, street trees, and sometimes private lawns, this paper focuses more narrowly on the use of vegetated strategies for managing stormwater runoff, sometimes explicitly called green stormwater infrastructure. Green stormwater management practices include a range of interventions that use greening either alone or in conjunction with highly engineered systems for the primary purpose of reducing stormwater runoff, either through promoting infiltration into groundwater or holding initial stormwater surges for later and slower release into the sewer system.

Promoting green infrastructure as a stormwater management technique in a city, instead of developing centralized infrastructure, necessitates greening everything from streets to school playgrounds and private roofs. Part of the rationale for adopting this approach in Philadelphia is that there are a series of indirect community and economic benefits that will be gained as green infrastructure is placed throughout the city. Indeed PWD's initial proposal included the results of a triple bottom line assessment as a way of selling the GI approach to community stakeholders and the EPA, asserting that the City's Green City, Clean Waters program would, in addition to preventing CSO events, provide economic benefits worth $500 million, social benefits worth $1.3 billion, and other environmental benefits worth $400 million to the city (The Philadelphia Water Department, 2009).

The triple bottom line assessment was based on several decades of research on urban greening and vegetation in which a wide range of community benefits have been identified. If we consider green infrastructure in its broader meaning to include not just stormwater management practices but all vegetated parts of the city including parks, trees, and lawn spaces, research has long shown considerable benefits for surrounding communities. These include environmental benefits associated with not only reduced stormwater runoff (Xiao et al., 1998) but also reductions in the urban heat island effect (Gaffin et al., 2008, Cao et al., 2010), and improved air and water quality (Nowak et al., 2006). Economic benefits include both increased property values near trees and greenspaces and increased sales along greened commercial corridors (Econsult Corporation, 2009, Heckert and Mennis, 2012). Social benefits include improved health for nearby neighbors (Coutts et al., 2010, Branas et al., 2011), reduced stress (Hansmann et al., 2007), reduced crime (Kuo and Sullivan, 2001), and promotion of social interaction (Coley et al., 1997, Kuo et al., 1998). Almost all benefits that accrue to nearby residents by virtue of living near green infrastructure would be expected to be applicable to green stormwater infrastructure, even though many of its common forms such as tree trenches, rain gardens, and planters cannot be directly used the way parks and greenways can. While there are some important caveats and these benefits have not always been universally demonstrated (Troy and Grove, 2008), as a whole they suggest significant potential benefits for GI which makes the equitable provision of GI an important concern.

The reality of these community benefits has led to the consideration of greening as an environmental justice (EJ) concern, and researchers have questioned the fairness of existing distributions of parks (Nicholls, 2001, Sister et al., 2009, Heckert, 2013), greenway trails (Lindsey et al., 2001), trees (Heynen et al., 2006, Landry and Chakraborty, 2009), and vegetation in general (Pham et al., 2012). The basic premise of these studies is that an inequitable distribution, especially when correlated with race or socioeconomic status, is a marker of environmental injustice. Attempts to measure the impact of greening programs have shown that they may reduce some inequities (Heckert, 2013), but also call into question the extent to which public programs can address differences when private property is involved (Heynen et al., 2006).

PWD plans to invest $1.6 billion over the first 20 years of the Green City, Clean Waters project (The Philadelphia Water Department, 2009). The fact that PWD is investing heavily in green infrastructure is fundamentally changing the nature of its relationship with communities. It is no longer a disconnected agency that just provides water, but is now one that is actively engaging in community development. PWD acknowledges that greening public land alone will not be adequate for reaching stormwater management goals, which call for “greening one third of the existing impervious cover in [the] Combined Sewer System drainage areas”, a goal they intend to reach “[w]ith the assistance of many public and private partners” (The Philadelphia Water Department, 2009, 4). PWD will need to coordinate with the multiple municipal departments, schools, non-profits, businesses, and private citizens to identify suitable land for GI installations and to see the installations through. This makes stormwater management a distributed system in which no single entity has full control over how or where implementation of GI occurs. Distributed systems provide new challenges for planning as power is shifted from PWD to other stakeholders. PWD has to engage more with the community and to make decisions about what types of projects should be developed where, with whom, and how (Travaline et al., 2015). This presents an opportunity for more inclusive and equitable decision-making about GI, but also presents a challenge in that the reliance on private property and private property owners to implement GI may inherently privilege residents with higher socio-economic status.

While the necessary inclusion of many additional voices opens the door for an inclusive and responsive approach to GI planning, we contend that this distributed nature of decision making also presents the very real possibility that GI investments might reinforce inequalities among communities. In a series of studies of tree canopy cover in Milwaukee, WI, Heynen, Perkins, and collaborators showed that inequalities in tree canopy cover were related to race and socioeconomic status, but that they were also based on tree cover on private, owner-occupied properties. This was characterized as a neoliberalization of nature that would be challenging for public programs to address because of the importance of private property use in producing the inequities (Perkins et al., 2004, Heynen and Perkins, 2005, Heynen et al., 2006). To the extent that PWD will rely on private property owners to install and maintain GI projects, these studies suggest that it may be challenging to promote GI in areas where residents are less likely to own their own properties, live on smaller properties with less available space for GI installations, or lack the financial resources to feel confident about long-term maintenance of the GI installations. While these challenges are not insurmountable, they are unlikely to be overcome if they are not explicitly considered in GI planning and communities with these challenges not targeted for additional support.

If this is the case, then how should PWD make decisions about where to site GI investments so that they are equitably distributed? Can we develop a GIS tool to help target investment so that it maximizes community benefits from GI investment while reducing disparities between neighborhoods? We argue that an equity index built specifically to evaluate existing conditions is an important tool to guide decision-making about GI. This would force policy-makers and citizens alike to compare neighborhood need and justify the distribution of GI investment in the CSO. The GI Equity Index builds off of a growing recognition that “bottom up” GIS technologies can be put in the hands of communities to help them use data to justify investment strategies that focus on filling “equity voids” (Talen, 2000). Such a tool may help shift the decision-making power to the community level as citizens start to understand how their communities rank against other communities in the CSO and where investment is going, raising questions about equity that might otherwise go unasked.

To promote more equitable distribution of GI investment, first it is important to define equity. While the term equity is used to mean fairness, there are several different approaches to defining what would make a distribution fair, and researchers have developed typologies of equity to get at these variations. Some of the differing definitions of equity are: equality, in which everyone has access to the same level of services; need, in which people or places with higher need receive more services; demand, in which people who express greater demand for services receive more; and market based equity in which service provision is related to either willingness to pay or actual differences in the amount that people pay for a service (Lucy, 1981, Crompton and Wicks, 1988, Talen, 1998).

Often in questioning the distribution of environmental amenities such as GI, equity is equated with equality. Wicks and Crompton (1986) found that both the general population and parks and recreation directors preferred equality over need, demand or market-based approaches as a method of allocating park resources. Similarly, the majority of EJ studies that have examined park, tree, and vegetation distributions have relied on equality as the indicator of equity, usually without any specific discussion of why equality is the appropriate benchmark (Boone et al., 2009, Landry and Chakraborty, 2009, Sister et al., 2009). In these studies, a distribution of services that shows differences correlated to race or class is seen as indicative of inequity compared to a null hypothesis of no difference. Lucy (1981) gives some indication of why this might be, noting that equality was often the standard used in early legal decisions regarding cases of environmental injustice over exposure to environmental risks. Nicholls (2001) provides an alternative approach in her exploration of equity in park access, in which she explicitly chooses a need-based definition of equity and argues that a park distribution that does not correlate to need would in fact be the inequitable one, while the truly equitable distribution would show a positive correlation with need.

Though popular, equality-as-equity may not be an ideal approach within the context of greening in general and PWD's program in particular since GI stormwater installations necessarily join an existing built environment and socioeconomic context, which would affect the community value provided by the GI installations. Recent studies of the impacts of greening on surrounding neighborhoods have shown that the impacts can vary based on neighborhood conditions (Troy and Grove, 2008, Escobedo and Nowak, 2009, Troy et al., 2012). The equality conceptualization of equity by definition does not take that community context into account so it cannot account for the possibility of different impacts in different areas nor is it likely to maximize overall community improvements.

The distributed management system in which PWD empowers communities to engage in GI planning while relying on private property owners to implement GI can instead be seen as inherently privileging the demand and market-based equity definitions. Community-based planning provides a forum for community-members to express opinions about GI, but in doing so it necessarily privileges those community members who express their opinions through attendance at community meetings, and may inadvertently ignore the concerns of residents who lack either the ability to participate or knowledge of how to do so. The reliance on private property for GI installation clearly privileges residents who own their own properties and disadvantages renters. To counter this, we argue for an explicitly need-based conceptualization of equity which acknowledges that communities with the highest need have the most potential to benefit from the program. While it is true that PWD will need to see GI implemented throughout the city if it is to achieve its goals of stormwater management and community development, not all stakeholders in all parts of the city will require the same amounts of assistance to implement GI practices and not all neighborhoods will accrue the same benefits from GI installation.

To promote equity using green infrastructure, we first need to determine which communities suffer from an “equity void” or “equity deficit”. By this, we refer to differences in the amenities that communities provide to their residents. Tiebout (1956) argued that communities offer a suite of amenities and residents will “vote with their feet” and choose neighborhoods that have a package of amenities that best suits their need and their family's needs. However, in low-income communities, residents are not able to “vote with their feet” because they often do not have the resources to move to more expensive neighborhoods that may provide more amenities. The result is that there are a large number of low-income residents who are “left” with communities that lack critical amenities that make a difference for quality of life. Researchers studying the impacts of living in disadvantaged communities that lack basic amenities have observed negative mental and physical health effects and long-term negative socio-economic consequences (Yen and Syme, 1999Leventhal and Brooks-Gunn, 2000, Ellen et al., 2001).

Given that not every community will have the same need for or capacity to implement GI, we argue for targeting GI investments and assistance to areas with the most need. By developing a GI Equity Index, we can systematically and transparently determine “need” and identify which communities are lacking community amenities that GI investment can help promote. We may also be able to create neighborhood typologies and test which GI practices are best suited for particular neighborhood contexts. For instance, in communities that lack parks, GI investment can be an opportunity to provide much needed recreational facilities. Efforts to green school playgrounds and provide recreational facilities may be important for communities with underperforming schools that also lack recreation space. Neighborhoods with high rates of unemployment may be places where the creation of green jobs associated with maintenance of GI might be very important. Here we should also note that an evaluation of the community capacity for a neighborhood to accept, plan for, promote, and maintain GI practices is critical.

Having settled on a need-based definition of equity, the next question is how to determine need. A strategy that has been increasingly been used by planners to measure equity and focus investment is the equity index, which combines a series of indicators to calculate a composite measure of need, deprivation, or risk. Much of the research on equity indices has been conducted in the field of public health and environmental justice. The Brandeis Child Opportunity Map (http://www.diversitydatakids.org/) is an index that examines children's opportunities as based on a combination of health, education, and other community factors. Another index is the EPA's new EJ Screen environmental justice mapping tool (http://www2.epa.gov/ejscreen), which incorporates not only data on groups that are particularly vulnerable to environmental hazards but also measures direct environmental risk such as lead paint exposure. The basic premise of these indices is that no single factor encompasses opportunity or risk, but that the true measure must take into account the many factors that influence it, enabling identification of areas that are particularly disadvantaged by being at risk based on several criteria.

There is a growing focus on using Geographic Information Systems (GIS) with indexes as a means of identifying opportunities for planning. The benefit of a GIS-based index is that it provides an easy, visual way for planners, policy-makers, and citizens to compare different neighborhoods according to a set of identified variables. It can also be used to promote more participatory planning processes. Talen's research on Bottom Up GIS (Talen, 2000) and Equity Mapping (Talen, 1998) highlights how GIS and mapping can help promote more equitable distribution of resources. A GIS-based index provides citizens and planners a shared understanding of what equity looks like in a community as it relates to need and public investment (Talen, 1998).

By developing a GI Equity Index for Philadelphia, we seek to demonstrate its potential applications in GI planning by asking first how the areas determined as highest need based on the GI Equity Index compare to areas based on a single measure, and second, whether or not there are areas identified as high need that do not currently have GI installations. We then use these applications as a starting point to discuss how use of an equity index in GI planning might help to ensure that disadvantaged communities are more empowered. The overall objective of our research is to show how an equity index can be used in the context of GI planning to promote equitable distributions of GI.