History and Philosophy of CLTs

The CLT model was inspired by older ideas of common ownership and different movements for the stewardship of land for wider community benefit around the world, including India, Israel, Mexico, Tanzania, and England (DeFilippis, Stromberg, and Williams 2018). In the United States, the CLT model is rooted in Henry George’s efforts at property reform and critique of the prevailing, individualist model of landownership, when civil rights activists merged George’s ideas with movements for community control. In 1969, the first CLT, New Communities, Inc., was founded in Georgia by a group of activists. New Communities was built on 5,000 acres of land, with the intention to decommodify land, provide for permanent community ownership of property, and give black farmers the opportunity to take and maintain control of basic resources.

New Communities was created as “a legal entity, a quasi-public body, chartered to hold land in stewardship for all mankind present and future while protecting the legitimate use-rights of its residents.” The community was conceptualized as a nested set of relationships, involving a resident community living within a trust, and a wider community of anyone who wanted or intended to be a resident or otherwise supported or identified with the trust. Ultimately, the CLT’s community included everyone who might support or benefit from the CLT’s mission, presence, and dramatic reorganization of landownership practices (DeFilippis, Stromberg, and Williams Reference DeFilippis, Stromberg and Williams2018).

The CLT movement reached urban areas in the 1980s as a way to combat increased property speculation in inner-city neighborhoods. Supporters felt that ongoing gentrification threatened the local community with mass displacement. In the 1990s, more favorable policies, increased funding opportunities, and shared learning across existing CLTs helped the CLT movement begin to thrive in the United States (Community Land Trust Network n.d.).

Long-Term Resource Stewardship in Residential Real Estate

Typically, the land stewarded by a CLT is used to build and provide permanently affordable housing into the future. The CLT owns the land but not the housing units on the land. Moderately low-income individuals can purchase a CLT house if they have a decent credit history and can pay for the home, which is typically restricted in cost to make it more affordable. The purchase is often funded through a special mortgage that applies to the house but not the land. Further, the buyer pays a small lease fee to the CLT, which stewards the property over the long term. In return, the homeowner is responsible for maintenance of the land.

When the current homeowner chooses to sell to the next qualifying buyer, the future homeowners are informed of the terms of CLT housing ownership. The selling owners receive the equity that they have paid into the home, as well as a portion (usually less than a third) of the increased value of their home. In this way, CLTs enable homeowners to build equity while maintaining a source of affordable housing that is made possible through the ground lease (Williams, Reference Williams2019a).

Grassroots organizations, activists, municipalities, and CDCs have all looked to CLTs as vehicles for promoting affordable housing and anti-displacement. Rising housing and land costs have exiled many families and devastated many communities, even in cities with the lowest unemployment rates (Williams Reference Williams2019a). CLTs are an effective mechanism when land and housing values are rising rapidly because they remove the speculative component of property values and create ownership opportunities for low-income people for the long term. Once it is part of a CLT portfolio, speculators and developers cannot purchase the land. The inability of real estate giants to develop property into lucrative hotels and condos creates a CLT neighborhood that is less likely to become gentrified and unaffordable to lower-income families. Further, the CLT-held land itself will certainly remain more affordable and accessible. Community organizers and other activists have become enthusiastic about promoting the CLT model (Hawkins-Simmons Reference Hawkins-Simmons2014).

The Dudley Neighbors CLT in Boston is a good example of this. Dudley Square is in the Roxbury neighborhood of South Boston and was a predominantly low-income area. Many vacant and tax-defaulted lots plagued the neighborhood, similar to Chicago’s Large Lots Program. Boston-area government and prominent financial institutions neglected the area as the neighborhood deteriorated throughout the 1970s. Over 20 percent of the land was vacant by the 1980s, with the remaining citizens composed of impoverished African Americans, Hispanics, and other minority groups that lacked the necessary financial resources to relocate. The City of Boston decided to transfer ownership of about fifteen acres, amounting to 1,300 lots, to a CLT.

Dudley Square residents responded by incorporating as a nonprofit called the Dudley Square Neighborhood Initiative (DSNI). They created a plan to build an urban village that would not displace the existing residents. This CLT built over 200 affordable homes on the vacant lots, a community greenhouse covering 10,000 square feet of land, a playground, an urban farm, and many other attractive amenities. CLTS collaborated with other community and grassroots organizations to implement the community’s vision for local development. They employed the CLT model to organize and construct a permanently affordable housing community that exercised the control of land necessary to fulfill their vision (Hawkins-Simmons and Axel-Lute Reference Hawkins-Simmons and Axel-Lute2015).

The Lincoln Land Institute of Land Policy has studied fifty-eight shared equity homeownership programs across 4,108 properties, where nearly three-quarters were managed by CLTs. Their analysis indicates that shared equity models have successfully delivered ownership of affordable homes to lower-income families across multiple generations. This is usually attributed to a combination of stable, below-market ground-lease prices and price caps on the resale of homes. The same study found that a shared equity model like CLT “provides financial security and mitigates risks for homeowners facing housing market turmoil,” and suggested that they helped low- and moderate-income residents navigate the volatile housing market fluctuations of 2008 more successfully than other conventional housing models (Wang et al. Reference Mironova2019).

Data has shown that such programs are serving buyers that might otherwise not be able to afford to purchase a home (discount metric) by preserving affordability for multiple generations of buyers (resale metric). Further, these programs allow homeowners to build wealth by retaining a portion of the gains in house appreciation or retiring mortgage debt. These shared equity programs offer a reliable path to sustainable homeownership. Grassroots organizations and CDCs have created CLTs in cities like Atlanta, Philadelphia, San Francisco, and Seattle to counteract urban gentrification (Mironova Reference Mironova2019).

Long-Term Resource Stewardship in Nonresidential Real Estate

Sometimes, CLTs steward land for both residential and nonresidential purposes through agriculture projects, green spaces, commercial spaces, playgrounds, other public centers for recreation or social services. For example, the Dudley Street (DSNI) CLT includes urban farm sites, parks and open space, and commercial properties (Loh Reference Loh2015).

Though less common, other CLTs steward land solely for nonresidential purposes. For instance, Southside CLT (SCLT) owns or directly manages twenty-one community gardens, manages another thirty-seven community gardens with partner organizations, owns or manages land used by twenty-five farmers to supply to various food businesses (restaurants, farmers’ markets, community-supported agriculture), and operates three production farms, across Rhode Island.

Local agencies, neighborhood churches, and private schools own and manage these gardens, and the SCLT assists them in providing both training and gardening resources. Their goal is to create community food systems for locally produced healthy food to be affordable and available to all, with a focus on serving low-income urban neighborhoods that may not have access to fresh produce. SCLT also aims to provide the people of Rhode Island with access to land, education, and resources to grow food in environmentally sustainable ways. For instance, they engage in youth employment and children’s summer learning programs, farmer training workshops and apprenticeships, hands-on training in food preparation and food growing. They match farmers with landowners who want to keep their land in production. Finally, they have advocated for increased investment in farmland access programs by the state of Rhode Island and founded the RI Food Policy Council to develop policies and partnerships to increase the capacity, visibility, and sustainability of the local food system (Southside Community Land Trust n.d.).

In another example, the Northeast Farmers of Color Land Trust (NEFOCLT) is a CLT as well as a conservation land trust. It was established to center the voices of and secure land for Black, Indigenous, and People of Color to farm, to conserve land by protecting native species ecosystems and engaging in regenerative farming and agroforestry, and to advocate for climate justice. NEFOCLT also collaborates with allied organizations to provide training and education in markets, business development, and financial planning. They have plans to acquire land to build a flagship community with incubator farms, commons for production, childcare, healthcare, and integrated ecosystem restoration (Farmers of Color Land Trust n.d.).

Nontraditional Property Rights

Some CLTs have attempted to use nontraditional property rights to retain community ownership over land. For example, Chinatown CLT (CCLT) has negotiated for ownership over easements (Lowe Reference Lowe2018). It has reclaimed municipal public land by cooperating with other community partners to convince the city government to undertake a disposition process for a public parcel of land, which is now set to become a multiuse development. This project encompasses 171 separate units of affordable housing. The effort was led by a local community development corporation collaborating with multiple for-profit partnerships and has received strong support from both community activists and residents who shared the project’s vision. It is an example of what can be developed by cooperative efforts between a local community development corporation and potential financial partners.

This is also an example of how a nonprofit CLT can assure public access to a given site without owning the land. It can do so by using an easement as a legal contract and entering in a deed. The CCLT now advises the community as it negotiates long-term public access through easements for an existing courtyard as well as a future branch library. This arrangement rectifies the broken promises given to residents who were promised access to public open spaces that were soon privatized into beautiful but privately gated courtyards. The Friends of the Chinatown Library will have secured an easement that ensures open public access to the library, and the CCLT will have an easement for open public access to a public courtyard.

CLTs might also fund or invest in other projects, where they do not own or lease the land. Sacramento CLT is working with other CLTs on a major project to turn a local creek and surrounding neighborhoods into a space that is walkable, bikeable, and safe. Here, they and other neighborhood groups involved receive funding and permission to act as stewards of public land (Sacramento Community Land Trust n.d.).

Community Governance

A critical component of a CLT’s mission is its efforts to promote a bottom-up style of democratic governance, which is a key component of land stewardship and community control over the land. Collaborative community governance is what gives historically marginalized people a way to participate in decision-making about how their neighborhoods are developed. Classic CLTs have two modes of community governance built into their organizational structures and bylaws. These were well-defined actions designed to keep the CLT aligned with the interests of its residents, and to be held accountable to both its residents and surrounding community or service area. This enables residents and those of the wider community (which could encompass a single neighborhood, multiple neighborhoods, town, city, or county) to influence critical decisions made by the CLT.

The first mode is having a governing board of directors that represents the interests of the people and locality that they serve. The governing board of a CLT has traditionally been “tripartite” in design. This means there is an equal number of seats for (1) homeowners leasing the land from the CLT, (2) surrounding community residents not living on CLT land, and (3) public and private sector stakeholders, such as public officials and nonprofit organizations providing assistance to the CLT or its homeowners. The governance structure of the typical CLT thus differs from the closed, private governance of other common-interest communities such as condos and cooperatives, since the boards of these entities consist only of private property owners (Wu and Foster Reference Wu and Foster2020). Board directors are typically volunteers, and their roles and responsibilities are set out in a CLT’s bylaws. They serve as key advisers on critical organizational questions, and often help with the foundational work of setting up the CLT when it is first starting out (Bath et al. Reference Bath, Girard, Ireland, Khan and Major2012).

Traditionally, directors are elected by the membership – one-third of the board is elected by members who live on the CLT’s land, another third is elected by general members who are part of the community but do not live on CLT-owned land, and the last third is elected by the total membership or the board itself to represent the public interest (Grounded Solutions Network 2018). For example, Dudley Square’s CLT board is represented by local and nearby residents who enforce lease agreements and other obligations of the lessors. The board is structured to reflect appropriate cultural or ethnic groupings that make up the Dudley community. The board has thirty-five seats, twenty of which are reserved for community residents. This group of twenty also comprises an equal number of representatives of the four main ethnic groups inside the community. Board members serve two-year terms, and are elected by the residents. The elected board vets and approves all decisions made by DSNI, but these decisions are always open to community input and participation (Wu and Foster Reference Wu and Foster2020).

Note, however, that the makeup and selection method of the board of directors may differ based on why and by whom the organization was created. CLTs may be established by an existing housing organization, nonprofit, or local government, as described earlier, and likely will not give CLT residents the same representation as a classic CLT. In these situations, the CLT’s board of directors may be partially or wholly appointed by the “parent” organization (which may itself be a membership organization elected by the community) (Grounded Solutions Network 2018).

The second mode is having a membership that elects the board of directors, which is similarly designed to hold the CLT accountable to its residents and the place-based community it belongs to.

The membership typically has two categories: one with all residents who live on CLT-owned land, and another with general members who opt in to support the CLT and pay annual membership fees. However, the makeup of a CLT’s membership varies between different organizations and their aims. For instance, they may retain the two categories but place restrictions on them. TRUST South LA, to achieve their mission of preserving opportunities for local working-class residents to remain in their neighborhoods, restricts their regular membership to low-income people who live or work in the land trust area. (T.R.U.S.T. South LA n.d.). Douglass CLT wanted a dedicated membership that is familiar with CLTs, which means their current membership is relatively small. While all residents may become lessee members, they are not automatically enrolled, but must affirmatively opt in. They have also put a cap on lessee members from different developments in their housing portfolio, to ensure that the membership is not dominated by voices from one development.Footnote ¹ Some CLTs may choose to add different membership classes – some organizations, such as SMASH, EBPREC, and Cooperation Jackson, have created a separate membership category for the staff of the organization.

At Douglass CLT, the key responsibilities of the membership involve attending the annual meeting and weighing in major decisions such as voting on the board of directors, bylaw changes or other changes to governing documents, or the sale of property. Members may also pay dues, review annual reports, or serve as board and/or special committee members. In addition to these, there are other opportunities for members to participate based on their skills or interests, such as organizing events or sourcing new properties (Douglass Community Land Trust n.d.).

Aside from these two iconic organizational structures, there are also other methods in which CLTs provide the community with methods for engagement, leadership, or control.

One such method is starting up specialty committees within the CLT, which means a wider group of interested members can learn and participate with more flexibility or based on specific interests. The San Francisco CLT has volunteer committees working in the areas of policy, membership, fundraising, finance, and projects (San Francisco Community Land Trust n.d.). Sacramento CLT has held working groups where members have convened around vacant public lands, tiny homes, and land acquisition strategies (Sacramento Community Land Trust n.d.).

Another method is participatory planning. Participatory planning involves engaging the community in what they want to see developed in addition to housing built on the land trust. This could be commercial projects, a greenhouse, or even developing agricultural land. For example, the priorities for Boston Chinatown’s development of the community land trust involved scouting out and developing a vision for remaining public parcels. This work involved engaging with communities about what they wanted to see there and how their community land trust’s priorities could be served.Footnote ²

A tangible example of how such decisions are worked out in the CCLT involves the Reggie Wong Park. The homes of one group of residents abut the park, in an area that includes the leather district. They wished to retain access to the park for family use while another group of residents wanted to play nine-man volleyball in the park. Even though some tension has existed around determining the top priorities for developing the park, the residents still have highlighted the opportunity to remain unified in the mission. The CLT brokered an agreement that allowed every group to access the park and helped build a bridge of understanding between the two groups, especially by communicating the Chinese residents’ stories. They helped publish an op-ed about why volleyball was so meaningful to the Chinese residents there. This work enabled these two groups to coexist in their use of the park.Footnote ³

Issues in Theory and Practice

While the CLT model has great potential to empower marginalized people and communities, and many CLTs in the past and present have successfully done so, it is important to acknowledge some limitations of CLTs today.

Academics have noted that the original goals for community ownership of land and local democracy may no longer be the priority for many CLTs today (DeFilippis, Stromberg, and Williams Reference DeFilippis, Stromberg and Williams2018). Rather, many practitioners and advocates have focused on aspects of the model most aligned with individual property ownership, and CLTs have shifted to act primarily as providers of affordable homes. The CLT acts as landowner, property manager, and steward of affordable units, with most of their staff devoted to housing development and stewardship (Hawkins-Simmons and Axel-Lute Reference Hawkins-Simmons and Axel-Lute2015). With affordable homeownership as the primary goal, CLTs may no longer see community control of land as the goal, and the provision of affordable housing as just one potential method of achieving this and serving the community.

Prominent CLT scholars have hypothesized that this shift is due to realities of funding, issues of scale, limited methods of successfully securing broader political support, industry perceptions of success, and limited staffing capacity (DeFilippis, Stromberg, and Williams Reference DeFilippis, Stromberg and Williams2018). Most CLTs are not self-sustaining, and are financially dependent on external funding sources, which come with their own stipulations and agendas (Mironova Reference Mironova2019). In recent years, the housing crisis and affordable homeownership has increasingly been prioritized on political agendas, with CLTs starting to gain traction as a model to achieve this. Many CLT funders have thus supported residential developments for this reason. Additionally, nonresidential developments are less lucrative and more logistically challenging (Williams Reference Williams2019b). It can also be easier to measure or demonstrate the impact of residential developments – for example, via the number of affordable homes secured. This may be why funding for nonresidential land uses, such as for playgrounds or community gardens, are much more difficult to find or attain.

Theoretically, the goals of affordable housing, reducing displacement, and community control go hand in hand, as they are all methods to assist low-income or marginalized people. However, academics have argued that the ground lease provides limited community control when CLT land is used for housing. For example, even though CLTs own land in perpetuity, community control is only exercised when new land is acquired. After the CLT leases the land out, typically for housing purposes, it is no longer able to make decisions, exercise control over, or access the land, as that is now the right of the homeowner to the exclusion of the rest of the community (Williams et al. Reference Williams, DeFilippis, Martin, Pierce, Kruger and Esfahani2018).

Furthermore, the focus on a narrow goal like providing affordable housing, rather than a broader, more systemic approach to empowering marginalized people through community control of land, means that many CLT funders may miss all the other ways the local community’s needs can be met or its residents engaged (Williams Reference Williams2019a). There is a myriad of ways that communities may need or want to use land beyond residential developments – urban agriculture, commercial development projects – which may serve different needs of local residents and aid the resiliency of the CLT itself. Indeed, one study has found that CLT involvement with urban agriculture or commercial development projects can drive economic development, diversify the CLT social base, build new partnerships, promote organizational resilience, and increase organizational visibility (Rosenberg and Yuen Reference Rosenberg and Yuen2013).

Another difficulty is that the burden largely falls on the frequently understaffed CLT to actively maintain a culture of participation. It can often be difficult to engage the community, as a study has found that efforts by staff to involve homeowners in greater participation within the land trust were unsuccessful (Thaden and Lowe Reference Thaden and Lowe2014). One study found that homeowners/people were drawn to CLTs because they needed a place to live, rather than to engage with the community (Kruger et al. Reference Kruger, DeFilippis, Williams, Esfahani, Martin and Pierce2020). In conversation with Douglass CLT, they found that it was difficult to show people that it was worth getting involved in. There is the constraint of time – residents have day jobs (potentially multiple jobs), childcare, and other responsibilities. They already interact with CLT staff for logistical practicalities. There is a second constraint of demonstrating to people that their efforts would have an impact. Participation can be particularly difficult with underrepresented people, who have previously felt used, or have not seen results from prior efforts in participation. Additionally, it may be intimidating for some to speak up or participate when they are put in a room with unfamiliar people, some of whom may have more assertive personalities.

Though these issues illustrate some challenges of the CLT structure and norms to achieve the goal of community landownership, CLTs have chosen this management structure to better serve local communities in ways that are not possible through traditional private landownership.

Interactions between CLTs

Between different CLTs, knowledge is often freely shared, and inter-CLT relationships are characterized by a collaborative ethic. Several CLT practitioners we interviewed mentioned that this is something they appreciate about the CLT sector and movement: practitioners are excited to share information about what they’re doing, and what has or has not worked for them.

The success and expansion of the CLT sector relies on documenting and publicizing useful data, and sharing technical knowledge, skills, and advice that covers the creation and operation of a successful CLT, between different CLTs through formal and informal networks.

These networks provide CLT members tactical and strategic support. The shared knowledge provides a starting point for practitioners to work from, despite the varying political and social landscapes that each organization occupies. They can speak about best practices for grant writing, strategies to collaborate more effectively with partner organizations, how to navigate property taxes, trends they are seeing, and to problem solve on making the CLT model work on a granular level. The network can help prevent duplication of effort between members to hopefully free up capacity for other work, and also leverage work that they each are doing to have maximum benefit. The regular communication and coordination also help local CLTs to align on their advocacy goals and shared reputation as CLTs to tell a larger, cohesive story across the region, which is important to secure funding, favorable policies, and continue to advance the narrative around CLTs.

Not only are CLT networks important spaces to think and problem-solve; the network and other members can provide solidarity and emotional support. As one practitioner said, it can sometimes be difficult not to get overwhelmed by the logistics and realities of running an organization or stewarding land. The networks and regular meetings can help remind them of their collective mission, principles, and the goal of collective community control and stewardship that is at the heart of what they’re trying to accomplish.

Within the United States, there are various networks and organizations dedicated to supporting existing and emerging CLTs through sharing experiences, connecting expertise, and deepening their knowledge base and capacities.

Interactions within CLTs

The next layer of interactions that we examine are the information flows that occur within any single CLT. For this section, we take a broad definition of what constitutes interactions within CLTs, including any interactions between formal members of a CLT (such as staff, board members, or residents), as well as with the broader place-based community of which the CLT is part.

While there has been extensive research on intraorganizational information-sharing flows, most of this literature has focused on the behaviors and sharing between the employees of an organization. This research has raised concerns around competition between employees, and lack of incentive to share information due to a company’s hierarchical structure and organizational culture (Tsai Reference Tsai2002). The CLT is a unique case study because almost none of the interactions we describe take place between CLT staff. This is because many CLTs have one or fewer full-time staff members (Wang and Rose 2018), and because the nature of a CLT job is external facing and requires intense communication with stakeholders outside their own organization. Rather, most interactions we have detailed take place between a CLT staff member and a stakeholder with different relationships, responsibilities, and viewpoints about the CLT. For example, a full-time CLT staff member is under an employment contract and paid for their work. On the other hand, board members have committed to responsibilities but are likely not financially compensated for that work, and most residents who are not in leadership positions have no obligation to participate in CLT governance at all. Many of the factors that influence quantity and quality of information sharing between employees of one organization therefore may not be relevant to CLTs or other NGOs that share similar circumstances. As such, the information flows we study within a CLT present new types of interactions that are less represented in information-sharing literature.

Next, we map out the common types of information flows within a CLT.

Smart City Planning and Governance: 2011–2016

In Philadelphia, the smart city imagined around 2011 was optimistic and familiar, essentially what was pitched in so many marketing materials and TED talks. Mayor Michael Nutter championed smart tech and the idea of Philadelphia becoming a smart city, both to solve local problems and to compete globally. In speeches, such as his 2012 keynote speech at IBM’s Smarter Cities Summit, Nutter incorporated “a common rhetorical theme in the … presentation of Philadelphia as a smart city, a theme enrolled in an ongoing, entrepreneurial economic growth agenda that oriented the city towards globalized enterprise” (Wiig Reference Wiig2016, 537). Themes of modernization and technological advancement permeated the rhetoric surrounding Philadelphia’s smart city initiatives. Creating a “[p]erception of Philadelphia as an innovative place with a dynamic economy was crucial to maintaining forward momentum to turn the city into a vibrant node in the globalized economy, instead of a failed, nineteenth and twentieth century industrial power” (Wiig Reference Wiig2016, 548).

The mayor’s ambitious agenda was exemplified by two distinct and very different agenda items (meso-level action arenas worthy of more detailed, independent study): the city’s engagement with IBM on the Digital On-Ramps project, and a series of Executive Orders issued by the Mayor’s Office.

Smart Tech Planning and Governance via Executive Order

Alongside the DOR project, Mayor Nutter adopted a pragmatic governance approach rooted in conventional public administration.Footnote ⁶ He issued a series of Executive Orders, which set goals, clarified responsibilities, provided direction, generated positions and working groups, and paved the way for smart tech deployment and data sharing in Philadelphia.

Mayor Nutter’s Executive OrdersFootnote ⁷ have their roots in EO 11-93 issued September 15, 1993 by Mayor Rendell. That order established the Mayor’s Office of Information Services (MOIS), which had the responsibilities of:

(1) establishing a strategic information technology plan, (2) establishing an IT steering committee to support strategic changes across the city, (3) to perform the needed assessment, development, operations, and maintenance of information systems applications across the city, and (4) establishing and operating an end user information center to provide re-engineering assistance and training in City standard IT policies.

The MOIS lasted until May 28, 2008, when Mayor Nutter signed EO 08-08, which established the Division of Technology (DOT). EO 08-08 focused on bringing Philadelphia into the digital era through the establishment of key positions and committees to oversee technology-based decisions. On July 6, 2009, Mayor Nutter signed EO 06-09, which effectively restated EO 08-08 with more explanation of various roles and responsibilities.

On August 22, 2011, Mayor Nutter signed EO 12-11, which took the existing DOT and restructured it into the Office of Innovation and Technology (OIT). OIT had the same responsibilities as the DOT, modified with language concerning interdepartmental cooperation and communication. EO 12-11 remains the guiding order on the structure and goals of OIT, which includes bridging the use of technology and information systems across city departments. EO 12-11 requires OIT to manage “[(1)] City telecommunications and information technology contracts, competitive bids, and requests for proposals … [(2)] establish information security policies … [and (3)] establish a Five-Year Strategic and Financial Plan defining … policies, procedures, processes, [and] guidelines … for Citywide telecommunications and information technology operation” (Executive Order 12-11, Information Technology, of August 22, 2011). The new OIT included several key positions and bodies including the Chief Innovation Officer (CIO), and the Technology Advisory Committee (TAC). The CIO serves as the head of the OIT and is responsible for managing the city’s technology and information systems as well as software acquisitions. The TAC serves to “provide strategic advice and counsel to the Office of Innovation and Technology and the CIO” (Executive Order 12-11, 4) It is comprised of thirteen people including the Managing Director, who serves as Philadelphia’s Chief Operating Officer, the Mayor’s Chief of Staff, the Finance Director, the Budget Director, the City Solicitor, the CIO, and several deputy mayors of various departments throughout the city.

Working in tandem with EO 12-11, chapter 21-2500Footnote ⁸ of the Philadelphia Code and Home Rule Charter requires that the Managing Director of the TAC produce and submit an annual information technology strategic plan to city council. According to the ordinance, the plan must “summarize and evaluate the current state of the City’s telecommunications and information technology infrastructure and detail and analyze the costs and benefits of the City’s plans for the acquisition, management, and use of telecommunications and information technology over the next five fiscal years” (Philadelphia Home Rule Charter § 21-2502).

On April 26, 2012, Mayor Nutter signed EO 1-12, which introduced five main goals for the City of Philadelphia concerning data use and transparency: the formation of an open data working group, the governance of data by an advisory board, an open government plan, an open data policy, and a social media policy. These five sections created policies and positions for administrators to do work involving data that has subsequently shaped Philadelphia’s data policies and plans over the last several years, including the SmartCityPHL Roadmap, which we discuss below. Notably, EO 1-12 established the Chief Data Officer (CDO), who is responsible for spearheading the open data initiative for the city.

EO 1-12 established an Open Data Policy in Philadelphia. The policy has eight main components, including the establishment of an Open Government Portal for sharing information related to the Executive Order, the establishment of an Open Data Catalog where each department in the city is required to catalog their data to determine if it needs to be shared in the Open Government Portal. The order also made clear the need for city departments to announce a timeline for the release of their data, the desire for public feedback as a mechanism for citizens to request data or share input about what information should be published, the identification of high-value datasets, and a six-month evaluation of the city’s progress toward fulfilling the open data initiative laid out in the Executive Order.

EO 1-12 directly led to the publishing of various governance documents articulating the city’s position on data collection and use as well as recommendations for how departments can share data either with each other or with the public. For example, as a result of EO 1-12, the City of Philadelphia published an Open Government Plan (GitHub 2013), which included themes of transparency, public participation, and collaboration.

The CIO and the CDO also published the Open Data Guidebook for city agencies. It describes best practices for preparing and sharing data across city departments and with the general public. The guidebook includes requirements for data structure to be compatible with the city’s metadata catalog, which is a repository for descriptive information concerning published datasets. The guidebook also supplies a methodology and reasoning for why publishing data is helpful for a city government:

By releasing open data, city departments may help to stimulate new and innovative ideas from our local technology community. There is great potential for open data to act as the fuel for new solutions and even new businesses that can address common problems or challenges facing those that live in, work in or travel to the City of Philadelphia.

(Headd Reference Headd2016, 2, 3)

EO 1-12 also led to collaboration with OpenDataPhilly, which is a private initiative managed by software firm Azavea. OpenDataPhilly is an online portal where people can request or publish datasets. The Open Data Guidebook specifies OpenDataPhilly as a place for city data to be published, after release by the CDO.

The Data Services Team within OIT provides a citywide data inventory online that describes every dataset that exists within city government and provides a mechanism for citizens to provide comments on whether the dataset should be released.

Not all datasets are released. Those that are released comply with standards set out in the Open Data Guidebook, which include anonymizing data, conducting a review, adding descriptive metadata, and ensuring data accuracy. As shown in the figures from the city’s metadata catalog website, many have been released on a range of different subjects (Figure 5.1).

Figure 5.1.

City of Philadelphia Metadata Catalog website

EO 12-11 and EO 1-12 stayed as the guiding force of Smart City initiatives in Philadelphia until February 4, 2019, when Mayor Kenney signed EO 2-19, The SmartCityPHL Initiative. That order formally established the SmartCityPHL initiative under the auspices of OIT, subject to the guidance of an advisory committee whose goal is to “guide and support the implementation of the key actions outlined in the SmartCityPHL Roadmap” (Kenney 2019, 2). OIT is responsible for maintaining the city’s open dataFootnote ⁹ programs and managing the SmartCityPHL initiative.

Smart City Planning and Governance: 2016–2021

By 2016, the aspirational smart city promoted by Mayor Nutter, IBM, and many others in 2011 seemed to have faded and been replaced with something quite different. Expectations and imaginations changed with experience. Pragmatism in the day-to-day operations of city government coupled with lessons in what smart tech and data could actually do (and not do) tempered exuberant optimism and hyped rhetoric. The DOR story was just one factor, among many. The shift in outlook and approach emerged gradually, beginning around 2014. While this observation surfaced in various interviews, we highlight the narrative of Ellen Hwang, who led the drafting process for the SmartCityPHL Roadmap and became the first Smart City director with OIT. In speaking with us, Ms. Hwang (2021) recalled her first job for the City of Philadelphia as part of the innovation management team in OIT:

Ms. Hwang (2021) explained a particularly influential experience (which another interviewee also mentioned). OIT had used the city’s “Small Contracts, Big Ideas” website to post Requests for Information (RFIs) as a “a good way for us to communicate things that government wants to learn about from folks who want to engage in procurement with the city.” One RFI concerned leased towers on several building sites that the city could be using more productively. Specifically, “thinking about … smart technology, … we cast a huge wide net, kept it super vague, and that RFI process, produced … over 200 responses.” Ms. Hwang (2021) then emphasized:

Ms. Hwang and her team convened interdepartmental meetings to “review these opportunities and to provide a strategic lens at seeing is there actually something bigger that we weren’t thinking about when we first set out to think about these tower leases?” She explains that mindset: “initially it was, how do we maximize assets that we have, tech assets that we have as a city for other things, and it ended up becoming what it is today … [the SmartCity PHL] initiative and having more cross-departmental dialogue around how do we actually adopt technology, and in a way that actually is meaningful and is actually going to improve city services.” During the conversation, she reiterated and emphasized the word actually, drawing a stark contrast with speculation and hype about smart tech. Yet the message was not pessimistic. Rather, as we discuss below, it resonated with the incremental, pragmatic, problem-solving approach to smart tech seen in the SmartCityPHL Roadmap.

In addition to improving interdepartmental coordination and communication, Ms. Hwang emphasized how the RFI-for-leased-towers experience demonstrated the need for the city to better leverage its resources, including limited budget, by being “more opportunistic about what is actually happening outside of city government.” This included building better relationships with industry and the community. She explained that the marketplace moved ahead of government and at the same time bombarded the government with proposals. “There was a lot of confusion on how to even just manage the sheer [number] of companies that were coming to city government.” To be more opportunistic and deal better with vendors and proposals, the community needed shared “terminology and education” about smart tech, and this required engagement and relationship-building – “everything that feeds into a good governance model.”

“From the beginning in 2016, the city approached the challenge strategically. Rather than tackle individual projects piecemeal, as so many cities have done, Philadelphia’s Office of Innovation and Technology (OIT) decided to create a roadmap that would guide and ensure long-term coordination of its wide-ranging projects” (Knowledge at Wharton 2018). Emily Yates, the current Smart City director for Philadelphia, described the roadmap as “a more organized approach to how we operate as a smart city and how we want to grow as a smart city.”

This shift to a strategic, holistic, city-wide plan required an assessment of what was already being done. OIT identified a variety of different projects, often siloed from each other. Figure 5.2, from the SmartCityPHL Roadmap, resulted from a thorough inventory of the city’s many existing smart tech assets and initiatives.

Figure 5.2.

City of Philadelphia SmartCityPHL Roadmap (p. 5), Survey of Existing Assets and Initiatives

Notably, many of these assets/initiatives constitute meso-level action arenas for which further GKC-based case study research would be useful. We discuss DataBridge in the next subsection and in our discussion of the meso-level action arena concerning Vacant Property Management.

To create a roadmap, the city established the SmartCityPHL initiative in early 2017, designated OIT as the lead department, and formed a working group including various city offices and departments. At the outset, the roadmap identifies four guiding principles – being locally inspired, innovative, equitable, and collaborative – and explains the methodology used to develop the roadmap – including an assessment of existing assets and initiatives, focus group interviews, identification of gaps and opportunities, working group brainstorming sessions, benchmarking against peer cities, and engagement with external stakeholders. The principles reflect a different set of community goals and objectives than had animated the smart tech vision in prior years.

The roadmap integrates smart city/tech planning and governance processes. It outlines three macro-level strategies: (1) building the policy foundation and governance infrastructure for meso-level projects and initiatives, (2) establishing consistent processes for engaging community members, and (3) developing sustainable funding mechanisms. We briefly describe the first two strategies, both because they relate most directly to the governance concerns of this case study and because they are developed in more detail in the roadmap; the discussion of funding is quite limited.

Smart City Foundations: Governance Structure, Policies, and Master Strategy

According to the SmartCityPHL Roadmap, “[a] strong governance structure is arguably the most important foundational element to smart city success” (City of Philadelphia 2019, 8). The roadmap establishes the following model: Executive leadership, which sets “vision and direction,” a Smart City director, who “collaborates and partners to inform and implement activities,” an Internal Working Group, which “collaborates and partners to inform and implement activities,” and an External Advisory Committee, which “provides advisory services, research, funding, solutions, and workforce capacity” (City of Philadelphia 2019, 8). In addition, according to Emily Yates (2021), the current Smart City director, “We are working to establish a task force, a third-party validation of sorts, so it will not be comprised of people in the Smart City advisory committee, it will be comprised of academics … to ensure that we are thinking of all community members.”

This structure, shown in Figure 5.3, aims to “drive … collaboration and standardization … [and to] solidify the City’s decision-making process, reporting structures, and roles and responsibilities of participating members.” While, at first glance, this new structure may seem to centralize decision-making and other governance powers within OIT, it does not go that far. Rather, it provides a central “hub” within city government for strategic development, coordination, communications, policy development and evaluation, and expertise regarding data, IT, and other related components of smart tech. Departments continue to have their own IT staff and make their own procurement, deployment, and other decisions regarding data and smart tech. The Smart City director and staff facilitate collaboration and function, at times, as consultants for other departments. During interviews, this role stood out for the city’s many different datasets (and also, to some extent, for network infrastructure, which we do not cover in this study).Footnote ¹³

Figure 5.3.

City of Philadelphia SmartCityPHL Roadmap (p. 8), Governance Structure

Hank Garie (2021), the city’s geographic information officer, explained that various programmatic or departmental datasets had been “messy” both for “historical” reasons and because IT professionals in different departments had been comfortable working in multiple databases without coordinating with one another.Footnote ¹⁴ This resulted in a lack of consistency in data standards. Other interviewees discussed the example of the water, streets, and fire departments possessing separate databases: “It makes it really difficult to figure out how to deploy services … you have electricity, and then you have a cable company, and … some … are public companies, and some … are private companies, and they’re all dealing with the same property but nobody can talk about the same property.” (We discuss this example further in the meso-level action arena section.)

Efforts by the OIT CityGeo team to address these types of concerns resulted in the creation of DataBridge, an integrated data warehouse and platform for geospatial and select non-geospatial data. DataBridge enables data sharing and provides city-wide infrastructure for analytics and applications.Footnote ¹⁵ OIT maintains the infrastructure, consisting of various pooled resources including technical know-how, software, and deployment managers. In alignment with the city’s general commitments to open data, transparency, and providing useful smart tech applications to the public, DataBridge aims to enable appropriate data flow to the city’s open data applications. Data flow enabled by DataBridge is shown in Figure 5.4.

Figure 5.4.

City of Philadelphia DataBridge

As seen in the figure, DataBridge enables different communities to publish and access data, and it also serves as the backbone for department-specific applications. DataBridge thus serves the administration, city departments, vendors, and partners that use the service to publish data, and it also serves the broader community that uses applications, such as OpenDataPhilly, OpenMaps, and CityAtlas.

Over the past ten years, the geographic information system (GIS) management team mostly focused on sharing data for “mapping, analysis, and city services.” In recent years, this department has focused on developing a secure and reliable foundation to coordinate, organize, and share city data internally and with the public. One interviewee suggested that the recent shift in the dialogue surrounding smart technology and data over the past two years has led departments to be “less skittish” about releasing data; it has become a “habit” for departments to publish their data. The interviewee explained that growth in literacy surrounding data publishing and the existence of mutual trust among stakeholders are catalysts in this culture change.

Coordination among stakeholders is critical to successful data sharing. For example, Kistine Carolan (2021) of OIT created a spreadsheet of all stakeholders to be informed and “kept in the loop” prior to the publication of any datasets. “[W]e’ve got a spreadsheet [of] all the stakeholders. And [it indicates] here’s the place in the process where we need to plug these people in, so that we make sure that when that data set hits the streets nobody’s caught by surprise.” At times, OIT may resolve conflicts among stakeholders. Hank Garie (2021) described a disagreement among the police department, the Mayor’s Office, and the press office about whether to share/publish crime data.

At the end of the day, the data out there is open data, and it’s been well received, and those people who had some concerns were able to back off and say “Okay for the public good, we’ll put out the data.” So it was very much a negotiated solution, but everybody who needed to be at the table had an opportunity to weigh in and express their concerns.

The regularly updated crime incident dataset is hosted on OpenDataPhilly alongside several visualizations for viewing and understanding the dataset. The dataset contains crime data from 2006 through to the present; each reported crime is categorized along with the block number where the incident occurred.

The SmartCityPHL Roadmap also recognizes that privacy and security policies are fundamental. The city has had privacy and security policies in place, as well as data-sharing agreements among departments, and the roadmap states high-level commitments to continue revising and updating those policies in light of changing circumstances and to “better understand the concerns of community members and privacy advocates when initiating smart city projects” (City of Philadelphia 2019, 9).

In addition, the roadmap calls for the development of a “formal master data strategy,” which, among other things, would include creating a formal inventory; establishing governance standards and technology specifications, and requirements for procurement, upgrades, and vendors; and workforce training. The master strategy would be guided by the following questions:

• What are we trying to understand?

• How will we manage sensor data?

• Who owns and manages the data?

• How will we secure the data?

• How will we analyze and operationalize the data?

Throughout our interviews, we heard that OIT has been and still is working on these policies and the master data strategy. Some interviewees noted that the pandemic had caused a shift in attention and priority and, as a result, progress on these issues had been delayed. (We do not delve further but note that this is another action arena ripe for further GKC-focused study.)

Community Challenges and Engagement

Beyond structure, policies, and strategy, the roadmap grounds the city’s approach to smart technology in pragmatism, which had been present and rising in prominence in prior years (as previously discussed). According to the roadmap, “it is important that any smart city project the City pursues is grounded in a challenge or problem that aligns with its broader mission and purpose.” To ensure such grounding and alignment, the roadmap outlines, as seen in Figure 5.5, a process for selecting and evaluating “appropriate and feasible smart city projects.”

Figure 5.5.

City of Philadelphia SmartCityPHL Roadmap (p. 16), decision-making process

The roadmap effectively outlines an applied version of comparative institutional analysis. “The first step is to establish a problem statement that conveys a challenge we are facing or trying to solve” (City of Philadelphia 2019, 17) and then “solution paths need to be identified” (City of Philadelphia 2019, 17). This step entails compiling a comprehensive list of ideas and potential solutions, drawing on internal and external sources. “Once we gather a long list of solutions, we will evaluate and prioritize them through criteria based on [a series of] questions” (City of Philadelphia 2019, 17). This approach highlights issues present in the city, and directly works to target initiatives toward solving the identified issues, seeking to achieve tangible improvements in the actual lives of community members.

Notably, the pragmatic process envisioned in the roadmap is a statement of strategy and does not constitute a binding rule and, to our knowledge, has not yet become a widely followed norm. Throughout the city, departments choosing to pursue smart tech projects need not follow these steps precisely. It remains to be seen whether/how the strategy and process become internalized.

The practical orientation to evaluating projects is further developed in the so-called Pitch + Pilot Program, usefully summarized in Figure 5.6 (from the roadmap).

Figure 5.6

City of Philadelphia SmartCityPHL Roadmap (p. 18), Pitch + Pilot

Pitch and Pilot, according to Emily Yates (2021, 13:40:48), is “[in] its very essence, a mechanism that was developed via the roadmap to create a more transparent way for the city to engage with the private sector to solve municipal challenges through tech and data.” It is a work in progress with a very limited budget, intended to be experimental and incremental. To date, the program has supported a handful of projects, summarized in Table 5.2.

Table 5.2.Pitch and Pilot programs

Finally, we note that the SmartCityPHL Roadmap places emphasis on engaging the community, through efforts such as forming subcommittees like the Citizens and Communities Subcommittee, which is “comprised of community-facing departments” and will be dedicated to “shaping and leading community engagement strategies and ensuring that solutions and projects will improve the lives of residents” (City of Philadelphia 2019, 8). In the second step of the governance process, identifying solution paths, the city will not only internally generate solutions, it will also “engage with a broader community to identify interesting and potentially impactful smart city solutions” (City of Philadelphia 2019, 16). Also among the factors utilized to evaluate and prioritize potential projects is community impact (City of Philadelphia 2019, 16). The criterion will ask whether “the program tangibly improves the experiences and outcomes of the community” (City of Philadelphia 2019, 16). Aspects of the roadmap such as these highlight the emphasis placed on community needs and involvement. Yet during our interviews, when we asked about how members of the broader community (beyond government) were actually involved in planning and decision-making processes, there was very little to report. Some interviewees acknowledged that there was less community engagement than promised or intended. The pandemic, we were told, made community outreach and engagement much more difficult. But, we were also told, it remains a pillar of the roadmap and the strategic plan, and it would be a priority going forward.Footnote ¹⁶

Background

Social Problems Caused by Vacant Properties

There are three types of social harms that arise from the vacant property crisis in Philadelphia (and elsewhere). First, there are effects on neighborhoods and residents, including:

• Crime: There is a relationship between crime hotspots in the city and areas with a significant level of vacant property (Philadelphia Land Bank Reference Lipp2015, 20). As described by Goldstein, Jensen, and Reiskin (2001, 2), vacant properties “provide convenient venues for criminal activity, such as drug trafficking and gangs.” Productively reusing these vacant properties can help to reduce neighborhood crime. For example, scholars at the University of Pennsylvania studied the effects of restoring blighted and vacant lots and among other results they found reductions in crime, gun violence and burglary (Branas et al. 2018, 2946).

• Safety: Community members – especially children and homeless persons – are often attracted to vacant properties, many of which have unsafe structures or other hazards. “Abandoned buildings attract the attention of neighborhood children, who may decide to use them as play areas, even though they are poorly maintained and may be unsafe” (Goldstein, Jensen, and Reiskin 2001, 2). In addition, “The poor condition of most abandoned buildings makes them fire hazards. In older neighborhoods with minimal space between buildings, fires can easily spread from one building to another. A fire in an abandoned building may very likely result in the destruction of occupied buildings, displacing residents and increasing neighborhood blight” (Goldstein, Jensen, and Reiskin 2001, 3).

• Trash: Vacant properties often become dump sites for trash and other debris (Loesch 2018). These trash-strewn sites not only attract vermin and are unappealing for neighborhood residents, but they are also costly for the city to maintain. “During my time at the City, we estimated that the city was spending almost $10 million a year cleaning up illegal dumping, oftentimes on vacant lots” (Esposito Reference Esposito2020).

Second, and related to the neighborhood issues, are impacts on the delivery of city services, including trash removal, policing, firefighting, and addressing water leaks on-site. Public service providers generally face an added degree of uncertainty in characterizing risks and providing services when the conditions of the property are unknown, causing administrative complexities and adding cost. “In weak markets characterized by rising levels of vacancy and foreclosures, tax revenue and property values fall and the cost of providing essential services like policing, fire protection, and code enforcement rise” (Philadelphia Land Bank 2015, 20).

Finally, vacant properties have significant economic impacts. Not surprisingly, vacant properties have a negative effect on the value of surrounding homes as well as perceptions of the neighborhood. “The average household loses over $8,000 in property value due to vacant property in Philadelphia” (Philadelphia Land Bank 2015, 18). As noted by Susan M. Wachter and Kevin C. Gillen (Reference Wachter and Gillen2006, 4) of the University of Pennsylvania, “Our findings indicate that adjacency to a neglected vacant lot subtracts 20% of value from a home relative to comparable homes farther away from the site.” In addition, abandoned or vacant property may also result in lost tax revenue for the city, which may impact the delivery of public services. According to a 2010 report by Econsult and the University of Pennsylvania for Philadelphia’s Redevelopment Authority, vacant property cost the city:

1. 1. “$3.6 billion in lost household wealth. Vacant parcels have a blighting effect on nearby properties, reducing values by 6.5 percent citywide …

2. 2. Over $20 million in city maintenance costs each year. Though the City controls only a fraction of the vacant parcels within the city, it has to bear significant costs to maintain all of them – waste clean‐up, pest control, police and fire – totaling over $20 million per year.

3. 3. At least $2 million in uncollected property taxes each year. 17,000 vacant parcels are tax delinquent, most by over a decade, owing a total of $70 million to the City and School District in back property taxes.” (Econsult Solutions 2010, v)

Where Vacant Property Management Meets Smart Technology: Meso-Level Action Arena

Vacant property management has always faced data and knowledge problems. Simply keeping track of the quantity and qualities of vacant properties can be daunting. However, such knowledge is essential to effective management (including sales) of properties and provision of public services. There also are significant challenges rooted in the distribution of knowledge and responsibilities concerning vacant properties. Over the past decade, Philadelphia has tried to tackle some of these obstacles by improving its organizational processes and procuring, developing, using, and sharing with the public data and smart tech. In this section, we describe those efforts and the roles that smart tech plays in vacant property management.

Knowledge Problems and Smart Tech Solutions

Given the significant impact of vacant properties and the needs of varied stakeholders, there are several knowledge questions that cities must answer. The following are representative:

• Location: Where are vacant properties located?

  • ◦ Are the properties clustered in specific neighborhoods?

  • ◦ What has been the impact of these vacant sites on the surrounding community?

• Ownership: Who has title to the vacant properties?

  • ◦ Are these owners in arrears on property taxes?

  • ◦ Which properties are owned by the city and which by private entities?

  • ◦ When and where should the city begin the tax foreclosure process?

  • ◦ Where and when should the city condemn and take the vacant property for repeated code violations?

• Status of the vacant site: Are vacant property owners maintaining the site, as required by Philadelphia law?

  • ◦ Are there crimes or safety hazards at the site?

  • ◦ What is the status of utilities, including water and electrical, at the property?

• Disposition: Which properties are for sale by the city?

  • ◦ What is the city’s plan for property reuse (commercial, multi-family residential, homeownership, side yard, community park, urban agriculture, etc.)?

  • ◦ Who and what are involved in the property disposition process?

  • ◦ How will upcoming property sales be communicated to the public, nonprofit community organizations, and developers?

There are many opportunities to use data, information and communications technologies, and smart technology to address these and related questions. In accordance with the macro-level strategy (discussed previously), the city has focused mostly on using the internet to provide usable open data as shared infrastructure for all stakeholders as well as mapping and visualization tools that leverage open datasets.

Smart tech deployment and governance in this action arena follows a familiar logic. The city has made a significant effort over the last several years to adopt technology policies that allow for more transparent government and enable open pipelines of data for everyone to use. Broadly speaking, the city sends data collected through various programs to DataBridge. Aggregated data is then published to a data repository, like the Philadelphia metadata catalog or a website like OpenDataPhilly. Once the data is published, it is available publicly for access and manipulation by anyone with an interest. The City of Philadelphia maintains a tool called Atlas for people to explore some of the published datasets.

The Vacant Properties Indicators Model (VPIM) dataset, owned and maintained by OIT, is an illustrative example of how data is published and utilized in the city. The dataset is hosted on OpenDataPhilly. Anyone can download it. The dataset is updated regularly by the city to ensure that the data present is the newest and most useful available. The city maintains two visualization applications that utilize the VPIM dataset: the Vacancy Viewer application and the Atlas application. Both are accessible either by direct link or through the OpenDataPhilly page for the VPIM dataset. On one hand, the Vacancy Viewer application is specifically tailored to the VPIM dataset. It serves to provide an easily navigable map-based visualization system for identifying and ranking vacant property according to the model created in the dataset. The Atlas application, on the other hand, serves a more general purpose. It helps visualize over fifty datasets. The Atlas application allows users to access information about properties across the city and allows users to learn about: property assessments, deeds, licenses and inspections, zoning, voting, and nearby 311 calls, crime incidents, zoning appeals, and vacant property. Atlas is an attempt by the city to unify visualization of geographically relevant datasets. The Atlas application contains a generalized visualization of the VPIM dataset to provide an entry point into viewing Philadelphia-centric data.

One of the stated goals for the DataBridge tool was to provide a streamlined methodology and process for publishing and viewing data. The Atlas application is an example of that streamlined process. However, as the Vacancy Viewer application shows, more specialized tools and visualizations can be created. One clear benefit of open data is innovation and creativity in the manipulation and viewing of the dataset. The City of Philadelphia has done a lot of work to create a platform and tools that can be used to observe and interact with the data they have published.

There are, in fact, many smart technology tools used to assess and manage vacant properties in Philadelphia. A sample of these tools are described in Table 5.3, sorted by the knowledge question each technology primarily addresses:

Table 5.3.Smart tools for vacant property management

Postindustrial Renewal and Economic Development

Pittsburgh’s first key dilemma consists of how to modernize an old, industrial city, with old material infrastructures; a declining population; an irregular geography; social and political infrastructures anchored in old institutions; many small neighborhoods disconnected from political power; formal fragmentation of government authority; and little reliance on modern data-focused systems. That dilemma includes day-to-day questions involving city living and working for residents and larger-scale questions involving how to grow and diversify the region’s economy, which is recovering from its former dependence on large-scale industrial manufacturing (Andes et al. Reference Andes, Horowitz, Helwig and Katz2017; Madison Reference Madison and Carpenter2012). Pittsburgh was an industrial city and region almost without peer. Today, Pittsburgh is unambiguously a postindustrial city and region. But the meaning and practice of its postindustrial status is in the process of being built – politically, economically, socially, and technologically. Economic renewal efforts still dominate the region’s political and cultural conversations roughly forty years after Pittsburgh’s steel industry collapsed.

The durability of the need and the difficulty of finding solutions testify to the depths to which Pittsburgh’s older industrial core shaped the region in every respect. It also testifies to the difficulty of marrying the legacy of that core to twenty-first-century technologies and governance. As Mayor Bill Peduto has said, smart city developments in Pittsburgh are linked closely to Pittsburgh’s emerging postindustrial identity, and the success of the new strategies depends on building on that core, not distinguishing “new” Pittsburgh from “old” Pittsburgh (Peduto Reference Peduto2015).

Public Administration

A second key dilemma involves the role of governance itself. Pittsburgh has experienced a conversion, from ideas of good governance as a means to the end of shaping Pittsburgh to good governance as an end in itself. The former perspective is highlighted by the public–private partnership embodied in the original, mid-twentieth-century Allegheny Conference for Community Development, described in greater detail later. The latter is highlighted by the idea of data-driven public decision-making as a modern value, embodied in particular in the contemporary Allegheny County Data Warehouse.

Not only have the aims of good governance and data-based decision-making changed, but as with all purported ideological shifts, practice may not match rhetoric, exposing social dilemmas within social dilemmas. After 2014, the City of Pittsburgh’s Department of Permits, Licensing, and Inspections was provided with digital technology for the first time with respect to many of its operations, both internal and public-facing. Snowplow operators and road repair crews were provided with tablet computers. Upgrades in the quality of service did not automatically follow. In part, legacy practices were simply difficult to dislodge, because incumbent staff members were comfortable with existing practices and were challenged by technology-based changes. In part, the material cost of technology outstripped the vision. The City of Pittsburgh circulated a call for proposals for smart street lights in 2018 relative to the city’s 40,000 street fixtures. It was imagined that the lights could be used for a mesh network of public Wi-Fi, would integrate with smart traffic control technology, and would monitor local air quality. The project was abandoned when city administrators realized that the effort would require installing thousands of miles of new network cables. Some obstacles are bureaucratic or logistical. Pooling data of different types and from different sources in a fragmented system presents considerable bureaucratic, labor, and technical challenges as data are generated to meet the details of different technical specifications.

Historically Grounded Inequities

A third central dilemma concerns the lack of alignment between Pittsburgh’s smart city goals and strategies with both community interests and research objectives at Pittsburgh’s key partners in nearby universities. As to the community, the problems that the City of Pittsburgh has tried to solve with smart city technology are not necessarily the most significant community-based problems that need to be addressed. As to research alignment, the priorities of Metro21: Smart Cities Institute at Carnegie Mellon University (CMU), which coordinates much of the relationship between CMU researchers and the City of Pittsburgh, are heavily influenced by partnerships between the institute and private industry.

Like many American cities, Pittsburgh suffers from profound inequities across different city neighborhoods and between the City of Pittsburgh and communities nearby, in Allegheny County and beyond, in the delivery of and access to basic amenities of urban living: public transit, education, clinical health and public health, clean air, clean water, safety and security, and economic opportunities. Smart city strategies were undertaken in part to begin to address those problems, by expanding the populations of citizens who were engaged in governance and community-level decision-making. Again, social dilemmas emerged within social dilemmas; historically excluded communities were skeptical of government solutions anchored in contemporary ICTs. In accessing government services, for example, people preferred to interact with human beings rather than with machines.

Political-Economic Hierarchy

Because Pittsburgh as a region is characterized by extreme formal fragmentation of political authority, overcoming obstacles and achieving coordination and cooperation among political organizations with respect to smart city practices is highly context-specific and often incomplete. Relevant mechanisms blend numerous formal and informal practices. In some smart city contexts, governance dilemmas focus on the privatization of public functions. That pattern is less pronounced in Pittsburgh. The relevant social dilemma focuses less on the role of private technology companies in dictating public policy and more on the ways in which public problems are solved by informal alliances of public, private, nonprofit, and philanthropic actors.

Some of the obstacles are budgetary. Until Mayor Peduto was inaugurated in 2014, the City of Pittsburgh Bureau of Police lacked any data analysts. Staffing has increased, modestly. Allegheny County, with greater financial resources and a significant track record in developing data analytics capabilities – funded initially by Pittsburgh philanthropy – provides voluntary data-related services and public-facing violent crime statistics dashboards for the City of Pittsburgh.

Some of the obstacles are jurisdictional and organizational. While Pittsburgh’s Western Pennsylvania Regional Data Center (WPRDC) is designated by both the City of Pittsburgh and Allegheny County as their official open data repository, the WPRDC has declined to accept and host certain datasets produced by the Allegheny County Data Warehouse, citing concerns that the Allegheny County data is not deidentified to the degree that the WPRDC and its other partners deem necessary. In 2021, the City of Pittsburgh launched a “Mobility as a Service” mobile application that integrates service data from the Port Authority of Allegheny County (an independent county-level entity that manages public transit services throughout the county, including the City of Pittsburgh) and private transit providers (technology companies offering ride-on-demand and carpooling services) with street-side access points and information hubs managed by the city.

Smart city strategies in these examples involve combinations of funding and relationship brokering that rely on third parties Neither the WPRDC nor the Allegheny County Data Warehouse would exist in their current forms today without substantial financial underwriting from Pittsburgh’s large philanthropic community. Pittsburgh’s Mobility as a Service initiative is funded by the Richard King Mellon Foundation. Many other smart city systems in Pittsburgh likewise rely on coordination among actors in the public sector and partners in Pittsburgh’s university community. That coordination is often multisided and therefore fragile.

Socioeconomic Hierarchy

Despite’s Pittsburgh governance fragmentation, historical wealth and technological expertise in Pittsburgh are highly concentrated in the region’s largest philanthropies and in its most significant research universities. Beyond those entities, Pittsburgh experiences extreme concentrations of informal cultural authority among political and business elites. Pittsburgh has long struggled as a community to access and distribute material resources effectively and equitably. It has also struggled to ensure appropriate and consistent levels of community participation in conversations about resource development and use. Smart city systems in Pittsburgh have been closely linked to the interests, expertise, and good will of a relatively narrow band of experts in addition to policy and institutional design.

Both the Allegheny County Data Warehouse and the region’s open data repository, the WPRDC, are strongly associated with specific individuals (Erin Dalton in the case of the Data Warehouse and Robert Gradeck in the case of the WPRDC) as well as with their commitments to good data practices in public administration. Like the design and operation of those organizations, collaborations between the City of Pittsburgh under Mayor Peduto and the Metro21 institute at CMU rely heavily on interpersonal relationships.

Those informal relationships mitigate the impacts of organizational polycentricity in part, because Pittsburgh’s interpersonal professional culture has long been noted for its collegiality. Professional and personal networks tend to be small and dense. CMU is not only a source of research for Pittsburgh’s smart city ventures. CMU’s degree programs are also the sources of graduates who have gone on to work on smart city practices in Pittsburgh, relying in part and building in part on a shared alumni identity. The University of Pittsburgh supplies not only a home for the WPRDC but also training and degrees and an informal alumni matrix for a number of Pittsburgh’s smart city actors.

Nevertheless, smart city practice in Pittsburgh is composed almost entirely of elite leadership with strong ties to local business, to national and international technology companies, and to smart city experts elsewhere. That pattern echoes (though it does not precisely replicate) Pittsburgh’s longstanding tradition of elite-led planning and strategy in both economic and cultural life. In a departure from that pattern, at times a reputation for smart city success in Pittsburgh has attracted expert talent from outside the region.

Informal relationships take on even greater importance as individual actors move from organization to organization and from role to role. They move both within Pittsburgh’s smart city ecology and also outside of it, establishing links with national smart cities organizations. Movement expands the pool of shared interpersonal expert relationships and helps to cement bridges among different smart cities organizations. Movement also potentially dilutes that pool, creating a new social dilemma. Even without movement, this informal network constructs bridges for expertise to transfer from organization to organization and sector to sector. That bridging also connects Pittsburgh’s smart city public sector and research communities to technology development practices in Pittsburgh’s private sector, including startup and spinout companies and Pittsburgh extensions of global technology firms.

Power Asymmetries: Democratic and/or Community Participation

The role of the Pittsburgh community as a whole in defining and shaping technology-informed governance has been relatively small. Pittsburgh’s smart city strategies have mostly been developed and deployed by the region’s political, business, and research-based elites, with little provision for community governance. The relative absence of broader community engagement is unsurprising in historical terms. Since the end of Pittsburgh’s steel industry, community distrust of newer technologies and their economic role has been a barrier to Pittsburgh’s overall renewal (Sabel Reference Sabel1993). CMU has a legendary research program in computer science and robotics, but that success has never translated into broad community-friendly sensibility. With respect to technology-related policy, community-based interventions in recent years have been sporadic. The City of Pittsburgh adopted a Dark Sky Lighting ordinance in 2021 largely as a product of community-based research and activism. The city’s Open Data Ordinance of 2014 likewise emerged in part from community interest. Both community efforts arose from engaged community volunteers rather than from broad, publicly supported outreach efforts.

Critical examination of the use of algorithms in public decision-making in Pittsburgh has come from the Pitt Cyber public policy program at the University of Pittsburgh (Pitt), likewise an initiative volunteered by expert community members rather than solicited by public authorities. That project is one of the few in Pittsburgh to recognize the significant misalignment between smart city program objectives and harmful community spillovers. In 2020, in response to an inquiry from Pitt Cyber, the City of Pittsburgh confirmed that it had discontinued a pilot predictive policing program, developed in partnership with CMU, called the “Crime Hot Spot Project.” Pittsburgh’s City Council followed that action with legislation banning police use of facial recognition technology without Council approval, although the Pittsburgh Bureau of Police later acknowledged using facial recognition technology (Clearview AI) during Black Lives Matters demonstrations in 2021.

The relatively small number of community-based interventions of that sort suggest that data collection and distribution practices may perpetuate rather than remedy inequitable living conditions in Pittsburgh with respect to health, wealth, and security both for individuals and for the community as a whole. Allegheny County’s Allegheny Family Screening Tool (AFST), a data-based system for allocating family support services, has been criticized on that basis, though more of the criticism has come from outside of the Pittsburgh region than from inside it (Eubanks Reference Eubanks2019). Smart city practices in Pittsburgh tend to consolidate rather than democratize control of Pittsburgh’s governance in the hands of political and business elites.

Even within Pittsburgh’s elite tier, the evolving strength of different voices is often difficult to discern. Elite leadership has gathered regularly in Pittsburgh to discuss strategies for economic development, though not specific to tackle smart technology issues. Decision-making, however, appears to be informal, consensus-based, and reliant on personal trust.

Given gaps between Pittsburgh’s smart city leadership and community participation, smart city technologies might be deployed to enhance community governance capabilities. Pittsburgh’s Burgh’s Eye View data dashboard project and other, similar data dashboards are nods in that direction. It is not certain that smart city designers are yet providing mechanisms for genuine community participation about smart technology governance in fair ways.

Instead, concerns about smart city technologies have been raised in the context of broader economic development decision-making rather than in the form of broad, direct objections to potentially harmful smart city practices. Incumbent Pittsburghers in some neighborhoods affected by technology-based economic development have protested the disruption of long-settled living patterns. In one well-known instance, the City of Pittsburgh backed the development of a technology-themed facility on a site adjacent to the Monongahela River that once housed a major steel mill, now called Hazelwood Green. Among the site’s amenities is a closed track for testing autonomous vehicles. Residents of the adjacent neighborhood, which lies between the Hazelwood neighborhood and the campuses of Pitt and CMU, strongly objected to the construction of a transit link that would connect the riverside site and the universities, the so-called Hazelwood Connector. They cited both the disruption of their neighborhood and the fact that the transit link would benefit only the technology elites. The dispute continues, sharpened by the fact that in late 2021, the University of Pittsburgh and the Richard King Mellon Foundation announced that the foundation was committing $100 million to help the university develop a biotechnology manufacturing facility at the site, provisionally named “BioForge.”

Information Asymmetries

Information asymmetries of various sorts mean that both acquiring too little data about Pittsburgh residents and too much data create opportunities for exploitation, corruption, and worse. I detected no evidence of bad faith or self-interested behavior in Pittsburgh’s smart city practices but abundant evidence of how Pittsburgh’s investments in partnerships with private high-technology companies and reliance on university-based research has skewed smart technology deployment so far. Residents may be unaware of political or historical conditions enabling data collection in certain domains and not enabling data collection in other domains. They may be led to believe that data collection and use is beneficial when in fact its impact is either neutral or possibly negative. Potentially harmful smart technology deployments may be difficult to detect and evaluate because robust mechanisms for transparency and oversight are not in place. That lack of salience or visibility not only limits residents’ ability to engage meaningfully in community-based or democratic oversight. It also limits their awareness of the extent to which smart city systems affect fellow residents and community members.

Information asymmetries may also reflect and generate dilemmas as to producing and sustaining social trust. As residents of a city anchored in neighborhoods and small communities of long standing, Pittsburghers traditionally exhibit high degrees of social trust in one another. That tradition does not always extend to trust in leadership. For historical reasons, some community members may be insufficiently trusting of relevant public and private leaders to engage in community-based governance of technology systems. Other community members may be too trusting of leadership and therefore may be uninterested in participating in collaborative governance efforts. Trust-based dilemmas of these sorts relate not only to trust in Pittsburgh’s leadership but also to trust (or lack thereof) among many Pittsburghers in technology itself, based on the region’s mixed history in building an economy on foundations anchored in twentieth-century industrial technology.

Managing Community Identity

A final social dilemma concerns the construction of community identity, both related to smart city technology use and in general. Community identity refers to how Pittsburgh and Pittsburghers see and represent themselves with respect to their history and their ambitions. The challenge is that not everyone in Pittsburgh participates in those conversations, let alone in the same way or on the same terms. Shared history and shared ambition are distributed unequally, as they almost always are in a given city. Yet there are important points of commonality. Building on that commonality is part and parcel of Pittsburgh’s smart technology practice. The City of Pittsburgh has tried to shape conversations about Pittsburghers’ community identity in the innovation economy, by trying to communicate to the broader public the effective and equitable public administration that can accompany public technology use. Beyond computing, smart city practices are linked to Pittsburgh’s efforts to reconstitute its public identity as an equitable and forward-looking “green” community in contrast to its older smoky self. Key actors blend advocacy and practice directed internally, to the Pittsburgh community itself, and persuasion directed externally, to political and economic development audiences outside of Pittsburgh. It is part of Pittsburgh’s smart city practice that Pittsburgh should see itself in smart, technology-based terms. It is also part of Pittsburgh’s smart city practice that others see Pittsburgh in those terms.

Shared city identity recapitulates additional social dilemmas. Both for historical and contemporary reasons, not all Pittsburghers experience or want to experience a shared “Pittsburgh” identity, whether related to technology use or otherwise. Promoting a collective, shared understanding of community identity may put at risk valuable ideas and behaviors as to spontaneity, serendipity, and personal development in both the experiences of residents and the behaviors of city planners, administrators, and public employees of all sorts. In contrast to cities such as New York and San Francisco that have long been celebrated for not only accepting but actively encouraging novelty and distinctiveness in human experience, Pittsburgh’s reputation lies at the opposite end of that spectrum. Generalizing, Pittsburgh is a place that encourages and sometimes even celebrates conformity and social stability (Madison Reference Madison and Carpenter2012). There are difficult but important balances to be struck between standardized, scripted, and even brittle behaviors in all elements of complex social systems, on the one hand, and improvised, innovative, and responsive behaviors on the other. Proponents of Pittsburgh’s prospective, novel postindustrial identity, including those who develop and deploy smart technologies, have to observe a poorly defined boundary between promoting shared community identity and pushing Pittsburgh residents in the direction of community rigidity and even inflexibility.

***

The summary of social dilemmas leaves important questions for further exploration and research. In what respects do the social dilemmas listed incorporate or point to subsidiary or overlapping social dilemmas? How should these dilemmas be characterized in terms of the tools, techniques, and concepts that are best used in elaborating their nuances and coming up with remedies? Are these urban planning challenges? Technology design challenges? Public administration challenges? Challenges regarding ideology, values, and purposes? All of these? How should observers and practitioners blend responses to questions of individual presence, identity, and activity with questions of collective, communal well-being? Documenting social dilemmas is only a beginning.

The next stage of this GKC-based investigation is describing the resources that have been implicated in smart technology systems in Pittsburgh; the actors involved in deploying, using, and overseeing those systems; and the roles that those people and organizations have played.

Data

“Data about Pittsburgh” consists of the first salient shared knowledge resource. Data includes data about Pittsburgh residents (including data about their interests, needs, and behaviors) and data about Pittsburgh as a physical place and space (including data about attributes of material infrastructures such as roads, lights, buildings, and parks). Future research may dig deeper into sector-specific and practice-specific data resources within this broad data domain. In a general sense and at both large and small scales, data about Pittsburgh capture the fact that quality of living in a communal context is a shared resource in a broad, fundamental sense.

Data that documents individual experience materializes that shared, aggregated resource and subjects it to new sorts of governance. The following sections describe extensive efforts by the City of Pittsburgh, Allegheny County, and other, related actors to collect, store, share, and use data across a broad range of smart city systems. As a shared resource, urban data is new in part, because of the novel technologies used to collect and manage it. Pittsburgh’s long history of collecting information about itself is documented in detail in the next section.

The various types of shared data in Pittsburgh include data derived from monitoring and observing environmental conditions (air and water quality; glare from street lighting; road damage) and human behavior (school attendance, movement of cars and buses). Conditions of data storage and use vary. Some are data stored in publicly managed systems for use by government actors (in particular, the Allegheny County Data Warehouse) and data stored in privately managed systems for use by both public and private actors (in particular, the WPRDC). Uses of the data vary widely as well. A lot of smart city data feeds into decision-making by government actors. Smart city data is formatted so that it is accessible and usable both by government actors and by residents and third parties, in particular, via the City of Pittsburgh’s Burgh’s Eye View data dashboards.

Expertise

A second shared knowledge resource is smart city expertise and expert governance itself, defined both by the positions and roles of decision-makers in relevant public, private, nonprofit, and higher education sectors and also by the substantive training, knowledge, and relational capital that individuals bring to bear on smart city practices. Both governance roles and the human beings who occupy them are subject to historical and political contingencies of numerous sorts. Shared expertise in Pittsburgh’s smart city context resembles shared expertise in many government and governance contexts, with the proviso that Pittsburgh’s industrial history has left a legacy of heavy reliance on locally developed and locally trained expert talent. Expertise in some cities is regularly and deeply refreshed by talented individuals moving in and out of the community, strengthening the social capital that underlies many effective city-specific and regional policy collaborations (Menashi Reference Menashi1997; Squazzoni Reference Squazzoni2009). For historical reasons, that has been much less common in Pittsburgh.

Even in Pittsburgh’s comparatively static setting, unlike data and datasets (which in principle can be documented as shared resources via organizational and technical criteria), expertise is a shared knowledge resource that defies simple description. Experts and expertise may be recognized by virtue of role, by credentials, by formal peer recognition, and/or by social acceptance in some relevant community (Hartelius Reference Hartelius2020). Expert networks are often fluid groups, and the expertise that they share is likewise dynamic. What counts as smart city expertise changes, as technology evolves, and as administrative and other governance strategies evolve (Eyal Reference Eyal2013). In Pittsburgh, smart city leaders and practitioners observe and learn from experiences in other places.

Taking those caveats into account, I observed a Pittsburgh-related “expertise community” for smart technology that includes substantial connections to CMU, both as a training ground for professionals in technology-based professions and as a key node in constructing research partnerships with industry related to smart city technology and practice; to Pitt, which has cultivated a node of similar type and function but which focuses less on industry partnerships and more on training professionals in public administration; to Pittsburgh’s philanthropic and nonprofit sectors, much of which are staffed by graduates of CMU and Pitt; and to the City of Pittsburgh and Allegheny County itself. People and their associated expertise circulate regularly within this network, moving from CMU to Pitt (Robert Gradeck, the director of the WPRDC, worked previously at CMU); from the nonprofit sector to CMU (Rick Stafford, the founding director of the Metro21: Smart Cities Institute at CMU, was previously the Executive Director of the nonprofit Allegheny Conference on Community Development); from Pitt to the City of Pittsburgh (Chris Belasco, Enterprise Project Manager at the city’s Department of Innovation and Performance, received his PhD from Pitt and serves as an adjunct professor there); and from the City of Pittsburgh to Allegheny County. Erin Dalton, Director of the Allegheny County Department of Human Services (DHS) and overseer of that department’s AFST system, holds a master’s degree from CMU.

These are tips of the proverbial iceberg. The number of personal and professional relationships evident in the construction of Pittsburgh’s smart city community is too numerous to document in full detail here.

Community Identity

A final, central shared knowledge resource in Pittsburgh’s smart city context consists of how individuals and small groups coalesce in time and over time to establish their collective identity as a city, producing both affective benefits and social trust that can underlie community development and improvement efforts (Sabel Reference Sabel1993). In short, Pittsburgh as an ideational construct is a critical shared resource, subject to social dilemmas as described earlier, that contributes to and follows from Pittsburgh’s smart city trajectory. A number of intersecting processes generate that construct. Political mechanisms exist for building and sustaining it, along with the dynamics of spatial relationships. Because the process isn’t coercive, some added ingredients are necessary. In significant respects cities are the durable products of processes of shared social cognition relative to everyday experience and relative to a place (Secor Reference Secor2004). Individuals signal their affective experiences to others in both purposeful and casual ways; they tap into histories of urban identity and shape its direction going forward. Key actors and nodes in cultural networks reinforce the salience of certain behaviors and cultural signifiers. In Pittsburgh’s technology practices, the largest local philanthropies have often performed this role, steering investment in smart technologies in ways that align with inherited understandings of the best interests of the community.

Taking account of the fact that these processes themselves are mostly immaterial, variable, and highly imprecise, in many cities in the twentieth and twenty-first centuries, including Pittsburgh, social cohesion and trust built on urban identity has been purchased by corporate interests. Business and political elites in Pittsburgh have repeatedly tried to capitalize on local research and development activity in the robotics sector by publicly promoting the idea that Pittsburgh has become “Roboburgh” (Dieterich-Ward Reference Dieterich-Ward2016).

Even more important, on a broad scale, have been corporate efforts associated with professional sports teams – American football, baseball, basketball, and ice hockey in the United States; ice hockey in Canada; football (soccer) in much of the rest of the world. In Pittsburgh, the shared community identity manufactured by workplace-based communities during the steel era (Slavishak Reference Slavishak2008) has long since been transformed into community affection for its professional sports teams, particularly the Pittsburgh Steelers American football team. Fanaticism in support of the Steelers is arguably the only phenomenon that unites most Pittsburgh residents as “Pittsburghers” across the region. Fan identity is materialized typically via the “Terrible Towel,” a small yellow terrycloth towel printed with a black “Terrible Towel” logo, that Pittsburgh residents and supporters of Pittsburgh professional sports teams twirl overhead while attending games in person, to celebrate and encourage Pittsburgh teams and fellow supporters. The Terrible Towel is an emblem and signal of Pittsburgh’s shared identity. Black and gold are the official colors of the City of Pittsburgh and the dominant colors of each of the city’s professional sports teams. They were part of the coat of arms of William Pitt, first Earl of Chatham, English Prime Minister in the late 1700s, for whom the city is named. Today, they form an integral part of Pittsburgh’s symbolic identity, together with the region’s steel history.

Actors and Roles in Action Arenas

Having sketched relevant shared resources and related social dilemmas, the next step suggested by the GKC framework is identifying how resources, actors, and their roles are assembled into “action arenas” or social contexts in which governance activity related to smart city technology takes place, generating outcomes. Taking account of public sector, philanthropic, and higher education institutions as key actors, smart city action arenas in Pittsburgh can be visualized in a general way as depicted in Figure 6.1.

Figure 6.1.

Smart city action arenas in Pittsburgh, Pennsylvania

Consistent with the discussion earlier in this section, the image represents Pittsburgh as a whole as an action arena. It shows both a series of subsidiary action arenas in the form of public sector entities, university entities, and philanthropic entities. It also identifies a distinct action arena that consists of actors anchored outside of Pittsburgh that engage in some respect with Pittsburgh smart city practices, including technology vendors, nonprofit organizations, and federal and state governments. The residents of Pittsburgh, both in themselves and in the form of community organizations and private sector companies, appear as constituent members of the macro Pittsburgh action arena. For-profit firms are not represented as an action arena in themselves, because I could not discern any evidence of collective or communal firm-based governance behavior in Pittsburgh, or with respect to Pittsburgh. Instead, both local private firms and national and international private firms interacted regularly with key actors in the primary government, university, and philanthropic sectors, selling technology and sometimes offering relevant expertise.

Two important considerations dictate relying only generally on the characterization represented in Figure 6.1, rather than too narrowly or precisely, in exploring smart city governance in Pittsburgh. First, each of the action arenas identified in Figure 6.1 signifies a number of smaller action arenas nested inside it. Each action arena, large and small, is subject to a greater or lesser degree to the resource descriptions and social dilemma characterizations supplied earlier. The overlapping circles speak to data itself as a resource and to governance expertise as a resource. For example, Pittsburgh governments include the City of Pittsburgh and, within the City of Pittsburgh, several distinct administrative departments. Pittsburgh universities include CMU and Pitt. CMU includes various research and other programs within CMU, such as Metro21, the CREATE Lab, and individual faculty members’ research programs. Likewise, Pitt includes various subsidiary units and researchers. Each of those should be considered an action arena with respect to smart city initiatives. Moving flexibly from larger to smaller scales in that regard is consistent with the intuition that smaller action arenas may be nested within larger ones.

Second, Figure 6.1 signifies that action arenas in the smart city setting are evidence of a polycentric social, cultural, and political system. Polycentricity highlights substantial overlaps in formal jurisdictional authority and in informal governance responsibilities. Yet that focus may detract from the fact that both formal and informal boundaries among action arenas often are less significant in Pittsburgh than interpersonal relationships among individual actors, including both social and political relationships. Smart city initiatives in Pittsburgh often require not only substantial collaboration among and across several polycentric centers but also among and across particular individuals, whose histories and forms of expertise accompany them as they migrate from organization to organization even within the Pittsburgh city action arena as a whole. In short, the important attention given to action arenas generally tends in Pittsburgh’s specific case to give insufficient weight to individual agency and to idiosyncrasies of personal history and attitude. As between governance system and structure, on the one hand, and personality on the other, a great deal of Pittsburgh’s smart city experience has been rooted in the latter.

Within these action arenas, judgments about how governance is produced are fluid. Smart city governance in Pittsburgh has not been heavily formalized by public actors. Formal, public law governing smart city activity in Pittsburgh is relatively modest in scope. A City of Pittsburgh ordinance passed in 2014 defines municipal obligations relative to publishing public-generated datasets in a publicly-accessible repository. Today, that repository is the privately supported and operated Western Pennsylvania Regional Data Center (WPRDC). Although the legal obligations and the creation of the WPRDC were part of a coordinated governance strategy for sharing data collected by the City of Pittsburgh and Allegheny County (with the operating costs of the WPRDC largely underwritten by leading local philanthropies), the repository was funded and launched only after the city subjected itself to a duty to share information, and as of late 2021 the city was not yet fully compliant with the law. Other City of Pittsburgh ordinances mandate certain private sector compliance and disclosure in connection with green construction and aspire to return “dark skies” to Pittsburgh via procurement and installation of improved streetlighting.

Informal rules, by contrast, govern most of this activity. The highest profile individual actors in Pittsburgh’s smart city ecology emphasize that in developing and deploying smart technology, they prioritize the interests of residents, both in term of how data is collected and used and in terms of acting consistently with principles of good government. There is no doubt that those views are genuine and motivated by good faith considerations. There is no doubt that views of what is possible and what is best are informed partly by deliberation about the future of Pittsburgh and about the future of good government generally, facilitated by conversations with colleagues in other places. There is also no doubt that these views are informed by knowledge about peer community practices and the uses of technology supplied by industry consultants and other third parties.

In practice, key Pittsburgh smart city actors invoke and rely on industry-standard practices regarding data security and data privacy. The City of Pittsburgh migrated its data storage architecture to Google Cloud starting in 2021, and the contract governing that commercial relationship emphasizes Google’s security practices. That contract does not specify undertakings by any party as to the privacy of residents or other data subjects. Interviews and document reviews for this study revealed no standard or typical practice by the City of Pittsburgh relative to sharing information with residents about possible privacy interests implicated by deploying smart city systems, other than consultations as needed with lawyers employed by the city and with third-party technical and policy experts. Nevertheless, the purpose of the move to Google Cloud is unambiguous: to use the Google Cloud infrastructure to build a “data lake” of pooled data for use in data analytics and data reporting. In homage to the specifics of Pittsburgh’s geography, and in contrast to Allegheny County’s Data Warehouse, the City of Pittsburgh pool is known as “Data Rivers.”

By contrast, public access to the WPRDC repository is governed, formally, by a click-through “Data Use Agreement.” That text is directed almost entirely to exonerating WPRDC and its sponsors and supporters from possible liability associated with using WPRDC-hosted data. As a practical matter, the WPRDC has no resources to follow up on or monitor compliance by community-based data users, and the disclaimer, like many click-through disclosures online, is both legally enforceable and, practically speaking, likely to be ignored. The presence of the disclosure does signify at least modest acknowledgment by WPRDC and its partners and sponsors that confidentiality, privacy, and security concerns are present when public data about resident activity is collected, curated, and shared. The WPRDC’s judgments about those values operate at a level that is tailored to its perception of its interests and those of city residents – as well as to the level of the University of Pittsburgh, which is WPRDC’s parent organization. Other actors express different judgments. Some datasets produced by the Allegheny County Department of Human Services have not been accepted for deposit with the WPRDC on account of differing understandings as to privacy protections afforded the subjects in the Allegheny County data.

The last notable feature of smart city governance in Pittsburgh, framed by the action arenas identified in Figure 6.1, is that smart city actors perceive that they are participants in a gift economy. That characterization applies both to their dealings with one another and also, at times, to their dealings with members of the broader Pittsburgh community. This is characterized partly in “pay it forward” terms, with the expectation that a kind of informal karmic justice associated with free and open sharing of civic data would eventually return benefits to the donor. It is characterized partly and more concretely in terms of overcoming obstacles to technology deployment by giving away time and expertise for free, particularly within large government organizations where time and technology expertise are not widely distributed in staff or budget terms. Representatives of technology firms that sell smart city technologies to cities distinguished their strategic consulting counsel as to smart technology uses from separate sales efforts. That perspective has both gift-oriented and profit-oriented motivations. The gift-oriented practices and attitudes confirm the existence of an informal network of favor exchange and loosely patterned cooperative behavior rather than a system of strong reciprocity or altruism (Jackson, Rodriguez-Barraquer, and Tan Reference Jackson, Rodriguez-Barraquer and Tan2012). The content of governance practices in Pittsburgh’s smart city contexts, or what might be termed Pittsburgh’s smart city “rules-in-use,” appears to be less significant for what they require or permit and more significant in that they confirm the existence of a community of smart city practice and expertise.

The Origins of Pittsburgh’s Intelligence

When it comes to smart city governance and to knowledge-sharing practices in particular, Pittsburgh is significant as much for who and what is left out as for who and what is included. Those patterns of inclusion and exclusion have deep roots.

During the 2013 campaign that led to his election as the sixtieth Mayor of Pittsburgh, Bill Peduto published a list of 100 actions his administration would initiate during its first 100 days. Number one on the list was “A 21st Century Pittsburgh Survey.” A Pittsburgh native and long-time member of the Pittsburgh City Council, Peduto brought with him a deep knowledge of the city’s history and a wish to see it achieve a twenty-first-century version of its twentieth-century glory. As mayor, Peduto aimed to replicate one of the first and greatest works of urban sociology ever produced for an American city.

The original Pittsburgh Survey, funded by the Russell Sage Foundation in New York and Chicago (then the Russell Sage Foundation for the Improvement of Living Conditions) and published between 1908 and 1914, appeared initially in thirty-five magazine articles and eventually was collected in six volumes of research (Greenwald and Anderson Reference Greenwald and Anderson1996). In its time, it was a first-of-its-kind, uniquely comprehensive data-focused examination of social welfare in an American community, synthesizing research on living conditions, working conditions, and industrial production in a single place (Lubove Reference Lubove1969).

Taking account of both the city itself and what the Survey, following common practice at the time, called the Pittsburgh Steel District, Pittsburgh was an enormously and almost incomprehensibly productive industrial place. During the nineteenth century it was known as the “Iron City.” During the twentieth century, the nickname was updated to the “Steel City.” The metallurgical metaphors were paired with a third, the “Smoky City,” due to Pittsburgh’s dirty air. An 1860s Pittsburgh travel writer noted that Pittsburgh was so vibrant with the fires and smoke of industry that he called Pittsburgh “hell with the lid taken off,” and he meant that as a compliment (Madison Reference Madison and Carpenter2012).

The “Steel District” geographic designation mattered to both researchers and local leaders more than a formal “City of Pittsburgh” identity, and related geography matters even today. Much of the steel production and associated industrial activity in Pittsburgh, including company towns, was located outside the City of Pittsburgh proper. Pittsburgh’s coal mines and steel mills were almost always located up Pittsburgh’s valleys, particularly up the Monongahela River and down the Ohio River, rather than in or near the urban center. The mills took advantage of the transportation economies that the rivers afforded relative to importing iron ore and exporting finished product.

Given the scale of the industry, workers associated with the steel industry – largely immigrants, in the late 1800s and early 1900s – were distributed around the Pittsburgh region. They were concentrated partly in company-supplied housing and partly in communities and neighborhoods adjacent to related industrial complexes, distributed across both the city and also in the less accessible, riverside locations that housed the largest mills. For most of its residents, the Pittsburgh Steel District was an awful place to live, with much of the population living in structures built to nonexistent housing codes and with virtually no modern water or sewer service.

Researchers for the Pittsburgh Survey aimed to document all of that, not to highlight anything specific to Pittsburgh but to use Pittsburgh as an exemplar of industrial conditions and social welfare across the United States. This was not, primarily, aimed at local reforms. The Survey was developed, researched, and written in response to an intervention by a small group of Pittsburgh business and community leaders who were aligned with the Progressive political movement nationwide. The vehicle for their interest was the Charities Publication Committee of New York; the host publication was Charities and the Commons: A Journal of Constructive Philanthropy. The point was fundamentally about Progressive politics: using data to support anti-corruption reform of public administration (out with patronage systems, in with the experts) and both voluntary and government intervention to improve residents’ social welfare.

(Pittsburgh wasn’t immune to ordinary efforts to improve urban living conditions. Around the same time that the Survey was researched and written, the City of Pittsburgh commissioned a report on its transportation infrastructure from a Chicago-based engineer. The report, released in 1910, was titled Report on the Pittsburgh Transportation Problem and criticized the lack of integration of the region’s many local streetcar companies.)

The Pittsburgh Survey generated massive amounts of data about industrial life, living conditions, and the environment. Locally, in practice, its impact was limited. (Pittsburgh had more success consolidating its streetcar operations.) To the extent that Pittsburgh absorbed the Survey’s lessons and welcomed political Progressivism, the movement took on a distinctly business-friendly character. The historian Roy Lubove chronicled in detail how the charitable impulses of the Pittsburgh business community married its market-dominating impulses during the 1920s, 1930s, and 1940s. Improvements were directed to physical infrastructure rather than to measurable changes to underlying questions of equity and social justice. Pittsburgh business leaders in partnership with Pittsburgh politicians endorsed and advanced housing reform legislation, investments in urban planning, and modest progress toward modern infrastructure, all in the interest of protecting and advancing Pittsburgh’s market positions in industrial production (Lubove Reference Lubove1969).

Throughout, the initiative to rely on the data and to begin the reforms depended on the essential political power of Pittsburgh’s business elite. That group consisted of a relatively small number of senior men serving as chief executives of large industrial firms that were, for all practical purposes, family-run enterprises. In the late 1800s and early 1900s, the group was led, politically, culturally, and economically, by Andrew Carnegie and Henry Clay Frick, industrialists, and Andrew Mellon, financier. Through the 1930s, their heirs and successors carried on the tradition of Pittsburgh leadership by Pittsburgh industry.

In 1943, informal collaborations between Pittsburgh’s business leaders and government leaders were consolidated and formalized in the Allegheny Conference on Community Development (ACCD), an early and durable public–private partnership in the form of a nonprofit corporation. The ACCD was energized largely by the leadership of the Mellon family banking and oil, gas, and coal concerns (embodied initially in Richard King Mellon), together with chief executives of other leading Pittsburgh companies (men named Mellon, Heinz, Kaufman, Hunt, and Hillman, and companies including Gulf Oil, Alcoa, U.S. Steel, Pittsburgh Plate Glass (now PPG Industries), and Heinz). This was noblesse oblige on the part of the individuals involved as much as corporate direction of state activity, in the guise of philanthropy. (The original by-laws of the ACCD required that member entities participate in the person of the company’s CEO or president, making elite governance formal and explicit.) The ACCD was made effective and durable by the active participation of Mayor David Lawrence (later Governor of Pennsylvania) and the local Democratic Party machine. The by-laws were later amended.

In the hands of the ACCD, in most respects, what needed to be done in Pittsburgh meant civic improvements to produce and reproduce the economic successes that defined the first half of Pittsburgh’s twentieth century. The ACCD took on the roles of coordinating regional planning across both business and local governments in the Pittsburgh region and of building community consensus around specific initiatives. In effect, the City of Pittsburgh and surrounding communities outsourced much of the visioning process to a public-spirited top tier of the private sector.

During the 1950s, the payoffs mostly consisted of productive investments in infrastructure: cleaning Pittsburgh’s smoky air by banning coal-fired home furnaces; cleaning the worst elements of Pittsburgh’s dirty rivers by regulating waste disposal; building modern highways and air transportation through Pittsburgh; organizing formal public health institutions; and redeveloping the most industrial sections of Pittsburgh’s Central Business District, replacing train sheds and related facilities with modern skyscrapers and parks. That initial round of improvements is often characterized by both historians and boosters as the “Pittsburgh Renaissance” (Madison Reference Madison and Carpenter2012).

During the 1960s, the payoffs mostly meant urban renewal, clearing out so-called slums (predominantly Black neighborhoods) and replacing them with amenities for Pittsburgh’s (predominantly white) professional class. The steel-domed Civic Auditorium, opened as a concert venue in 1961 (and demolished fifty years later, having acquired an afterlife as a sports arena), was intended to showcase Pittsburgh’s metals industry for the benefit of prospective investors in the region. Its construction eradicated the much of the cultural center of Black life in Pittsburgh, known as the lower Hill neighborhood, home to a thriving arts community and to more than 8,000 people.

Pittsburgh’s Intelligences of the Late Twentieth and Twenty-First Century

That rhythm – a data-fueled baseline for good government and welfare improvements, followed by an elite-driven, intuition-based, largely privatized set of visions, strategies, and tactics – defined Pittsburgh for much of the twentieth century. The pattern can be documented and illustrated further with efforts by Pittsburgh public authorities – and iconoclasts.

Around the same time that the Pittsburgh Survey was being produced and published, Frederick Law Olmsted, famous as one of the designers of New York’s Central Park and other well-known public parks and recreation facilities in the United States, was retained by Pittsburgh’s business elite to produce plans for Pittsburgh. As planners and designers, Olmsted and his firm would rely on the bureaucracy of urban planning to implement their visions, but their charge was to tame Pittsburgh’s appalling physical and social conditions and, in cleaning and regularizing the conditions of urban life, to instill the working people of the city with “appropriate” moral order (Ingham Reference Ingham1991). This was Progressivism at work in a different register, top-down rather than, as with the Pittsburgh Survey, data-driven and bottom-up. As in the work of Ebenezer Howard (author of the utopian planning guide Garden Cities of To-morrow: A Peaceful Path to Real Reform in 1898), orderly and systematic urban planning – in a manner of speaking, the smart city of yesteryear – was a mode of social reform (Beevers Reference Beevers1988).

Olmsted’s vision, delivered in a report in 1910, was adopted only in part. As with the results of the Survey, pragmatic physical improvements were pursued while social justice implications were ignored. Today, many of Pittsburgh’s larger boulevards and bridges owe their origin to Olmstead (Bauman and Muller Reference Bauman and Muller2006). Pittsburgh’s regional parks and nearby vacation destinations were built in the same era, displacing local working class communities in the interest of the patronizing impulses of the business community (Dieterich-Ward Reference Dieterich-Ward2016).

The top-down planning impulses of Pittsburgh’s power structure had additional manifestations, with a more entrepreneurial character. Pittsburgh’s most celebrated work of modern architecture, the Frank Lloyd Wright masterpiece Fallingwater (a vacation house located in the Allegheny Mountains just southeast of the city), was commissioned by Edgar Kaufmann, Sr., a local department store magnate. Kaufmann was so taken with Wright that he commissioned the architect in the 1940s to produce a series of futuristic plans for a civic center and related infrastructure to be built at the Point, the tip of the Downtown Central Business District. The civic center never came to pass; there is little evidence that the plans were ever seriously considered by the city. Kaufmann’s instincts were on the right path, however. The Point was leveled and remade as part of the Pittsburgh Renaissance during the 1950s. Among the new, related developments was a series of high-rise cruciform buildings clad in chrome-alloyed steel that evoke the 1920s Radiant City “Towers in the Park” vision of modernist, technocratic urban planning promoted by the architect Le Corbusier.

The planning impulse did not abate. In 1963, researchers at the University of Pittsburgh, together with the Pittsburgh Regional Planning Association (a subsidiary of the ACCD), produced a three-volume study addressing the economic prospects for the region, titled Economic Study of the Pittsburgh Region (Chinitz Reference Chinitz1961; Lubove Reference Lubove1965). It concluded that Pittsburgh’s economy was stagnating as the era of structural steel production in the area was likely coming to an end. The study called for a transition to a more technology-driven economy. That recommendation was all but ignored. A complementary effort to develop a comprehensive computer simulation of Pittsburgh’s land-based resources to support data-driven planning efforts was terminated and abandoned before it could be completed (Brewer Reference Brewer1973).

At the other end of the spectrum of community institutions, federal antipoverty programs and model cities initiatives during the 1960s encouraged the development of neighborhood-specific organizing in Pittsburgh. Those organizing efforts included the preparation of a Pittsburgh Neighborhood Atlas during the early 1970s, which surveyed residents about neighborhood satisfaction and satisfaction with public services, and documented data from seventy-eight Pittsburgh neighborhoods about real estate prices, loans and tax delinquencies, and welfare assistance. Funded largely by the University of Pittsburgh through its School of Social Work and completed in 1977, the Atlas and the organizing behind it contributed significantly to defining Pittsburgh’s neighborhoods in their modern configuration (now ninety in all) and to cultivating the neighborhood – otherwise omitted from the vision advanced by the Allegheny Conference – as an effective locus for community participation in planning, reconstruction, and economic development (Cunningham et al. Reference Cunningham, Ahlbrandt, Jewell and Hendrickson1976; Lubove Reference Lubove1996).

The Atlas was conceived and produced specifically as a counterpoint project, contrasting common perspectives against elite perspectives, rather than as a complement to the efforts of regional leaders. Following the collapse of the steel industry in Pittsburgh in the early 1980s, Pittsburgh leaders again promoted efforts to anchor the region in integrated visions of technology-based industry. In 1985, a coalition of Pittsburgh leaders (the ACCD, the presidents of the University of Pittsburgh and Carnegie Mellon University (CMU), the Mayor of the City of Pittsburgh, and the political leaders of Allegheny County) published the Strategy 21 report, proposing an economic development plan for the region in the wake of the end of the steel era (Deitrick and Briem Reference Deitrick and Briem2021). The report recommended pursuing an elaborate, data-focused effort to diversify the region’s economy away from its historical reliance on heavy manufacturing. Only one of the report’s significant recommendations was adopted. Pittsburgh built a major new international airport.

Out of the Furnace and Toward the Smart City

Pittsburgh’s historical tension between empiricism and elitism offers the key byway into understanding Pittsburgh’s smart city conditions today, even as the specific shape of that tension changed. It has been suggested that the rise of Pittsburgh’s two leading research universities and its largest philanthropies during the latter half of the twentieth century, and the decline of Pittsburgh’s old industrial and financial sector firms, brought with it a loss of interest in the cultural fabric of the city. The old industrialists’ possibly patronizing but nonetheless real focus on the lives of the people disappeared in favor of a focus on metrics (Lubove Reference Lubove1996). The criticism is overstated. In practice, governance technologies and tactics changed, and with new tactics came new goals. Good governance became measurable, at least in principle, rather than simply evident in residents’ and companies’ experience. The new players emerging in the later twentieth century and early twenty-first century brought forward a new and explicit focus on public administration and governance as goals in themselves, sharpening a distinction between this more modern, technocratic attitude and the noblesse oblige that inspired the original ACCD.

In other words, the era of a small number of supremely wealthy families in Pittsburgh actively driving the direction of the city, as heirs to the industrial and financial leaders of the late nineteenth century, ended. New key players emerged, taking their places alongside political leadership and the leaders of the region’s largest private companies, particularly the leaders of the foremost universities in Pittsburgh; the leaders of its largest employer; and the leaders of its major philanthropic organizations.

In 2013 and early 2014, Mayor Peduto’s list of 100 inaugural actions did not include turning Pittsburgh into a smart city or producing a smart city strategy. The list of 100 actions did include a number of items that fall within anyone’s definition of smart city administration. More important than the list itself, however, the Peduto administration helped to consolidate preexisting Pittsburgh assets and investments in data-driven government and private sector technology development, and to accelerate Pittsburgh’s reliance on smart city systems by weaving narratives that expressed smart city visions. This subsection summarizes the assets first, and then the visions.

The Blossoming of Pittsburgh as a Smart City

In addition to these relatively grand plans, Pittsburgh’s contemporary smart city investments have a variety of recent specific antecedents in both public and private sector technology deployments. Pittsburgh has never tried meaningfully to follow a plan regarding information technology, either as mode of government practice or as focus of economic development (Deitrick and Briem Reference Deitrick and Briem2021). Its efforts have advanced on the ground at lower levels. Some present practices are traceable to initiatives from twenty years ago and before. The Pittsburgh Neighborhood Atlas, from 1977, was a significant early modern effort to build an information system for the city. Other key early smart city ventures are highlighted here.

Three of these focused on civic data and public administration in the City of Pittsburgh. The first, 3 Rivers Connect (3RC) (named for Pittsburgh’s location at the confluence of three rivers), was a private nonprofit initiative launched in 1999, founded and operated by researchers connected to CMU, leveraging privately developed, venture capital-backed database, search, and visualization technology distributed in a pair of sister companies, MAYA Design and MAYA Viz. It was funded by Pittsburgh-based philanthropy. Characterizing itself as a venture in “civic computing,” 3RC initially hosted a web-based resource titled the Information Commons, which consisted of an early online directory of community-based organizations and resources. The Information Commons evolved into an effort to develop search tools and data analytics that crossed traditional and fragmented data silos, linking information from and for public sector organizations, economic development interests, and community groups.

(Notably, this early investment in the immaterial, technocratic city emerged around the same time that the early internet materialized in a physical location. In 2002, the ground floor of the former Downtown headquarters of Alcoa was converted by a public–private regional government entity (the Pittsburgh Regional Alliance (PRA)) into a “technology information hub” called the Xplorion. The Xplorion featured banks of plasma screens displaying information for visitors about Pittsburgh-specific business development, education and training opportunities, and cultural attractions. Before the smart city was conceived as an immanent part of everyday life, a version of the smart city in Pittsburgh was a showroom that one could visit on foot, and that the PRA hoped would create a “wow” factor that would appeal to businesses considering whether to locate in Pittsburgh.)

As a use case for MAYA’s technologies, 3RC itself grew in scope over fifteen years of operation, and it eventually developed and offered separate websites and software tools for both community and public sector application. 3RC not only inventoried resources across multiple sectors but also supplied tools for querying and analyzing data pools. Among its public sector partners was the Allegheny County Department of Human Services. The county’s humanservices.net domain served as a gateway to an Information Commons repository of information about daycare centers, drug and alcohol assistance, and food banks, combined with mobility and access information for citizens. As commercial search technologies and accessible databases got larger and more powerful, the case for 3RC weakened, and it wound down in 2012. Its privately owned data analytics technology for civic infrastructure had already been spun forward into a separate commercial entity.

The second was implemented within the administration of the City of Pittsburgh during the tenure of Mayor Tom Murphy in the early 2000s. The system was CitiStat, a statistics tracking and data analytics system pioneered in the City of Baltimore, Maryland. The purpose of CitiStat was to centralize data collection as to forms of citizen/community interaction (citizen phone calls about city services, pothole filling, garbage collection, and so on) and then to allocate service-based resources accordingly. The system would help rationalize the distribution of those resources and create a data-based system for employee and managerial accountability. Pittsburgh’s CitiStat system required a dedicated physical space where team members would meet to share and analyze data. The space was built, but the system did not survive the end of Mayor Murphy’s administration in 2006.

The third, serving most directly as a precursor to contemporary smart city practice, was the Pittsburgh Neighborhood and Community Information System (PNCIS), which operated from 2005 to 2014 as an initiative of the Center for Economic Development at CMU and the University Center for Social and Urban Research (UCSUR) at Pitt. Beginning in 2008, PNCIS was as an affiliate of the National Neighborhood Indicators Project (NNIP), a national network of data intermediaries organized by the Urban Institute. PNCIS was an open data initiative, collecting and cleaning datasets addressing public and community activities in Pittsburgh and making them available for public access and use, either online or via hard computer media. It was funded partly by Pitt, partly by CMU, and partly by the City of Pittsburgh, with fundraising and management support from a Pittsburgh-based community financial organization, the Pittsburgh Partnership for Neighborhood Development (PPND) and local philanthropies. But the PNCIS was not funded sufficiently for its services to meet the full range of community need, and it was not charged with supporting public sector activities as well as community organizations.

The technical and structural limitations of PNCIS were recognized and addressed in the development of a successor open data enterprise, the Western Pennsylvania Regional Data Center (WPRDC), which is supervised by the same person who led the PNCIS, Robert Gradeck. The WPRDC is the official open data repository of the City of Pittsburgh and accepts open datasets from all manner of regional governments and community organizations. It was created in 2014 via a collaboration among the City of Pittsburgh, Allegheny County, and local philanthropies. As noted earlier, a complementary open data ordinance was adopted by the City of Pittsburgh at the same time.

The impulse to collect and publish data as “indicators” of community well-being both preceded the WPRDC (and the PNCIS) and survives it. The Pittsburgh TODAY Regional Indicators project, housed separately at UCSUR, traces its origins to the mid-1990s and a regional benchmarking project initiated by the Pittsburgh Post-Gazette, the region’s principal daily newspaper. The project originated in the newspaper’s instinct to document the region’s post-steel recovery in quantitative terms. The project has been sustained through organizational and funding changes, including management by 3RC, and continues today under the leadership of one of the journalists who helped launch the project in the first place. Public sector indicators projects have been less durable. A Pittsburgh Equity Indicators report was published by the City of Pittsburgh in 2018 and updated in 2019, describing economic conditions in Pittsburgh relative to gender, race, and income.

Two additional enduring early smart city investments turn up at Allegheny County, home of the Department of Human Services mentioned earlier, and in the private real estate development community.

Allegheny County created its Data Warehouse in 1999, consolidating its own internal data relating to behavioral health, child welfare, and homeless services in order to support decision-making, improve case management, and conduct policy analysis. The Data Warehouse is, in sum, an internal management tool. (The 3RC service was in part an early public-facing interface for certain data collected in the Data Warehouse.) Later, data from other county agencies were included, and the county crossed jurisdictional lines to partner with the Pittsburgh Public Schools, an unrelated public authority responsible for all public primary and secondary education in the City of Pittsburgh; with the Allegheny County court system and the Allegheny County Jail; and the housing authorities of both Allegheny County and the City of Pittsburgh. Cooperative agreements with other government organizations enable the county to conduct trend-based data analysis that links county-level data to human services data acquired both from the state of Pennsylvania and from the federal government.

The Data Warehouse was launched as part of a larger, comprehensive reorganization of the county’s service departments, whose fragmented character was deemed to have contributed to the death of a child formerly in the charge of the county’s child protection service. Funding for the project came from Pittsburgh-based philanthropies. In expanded and modified form, it is still in use today. In 2013, with data shared by the Pittsburgh Public Schools and other school districts in Allegheny County, and in coordination with the United Way of Southwestern Pennsylvania, Allegheny County launched a “Be There” campaign addressed to public school students, premised on data-derived correlations between school attendance and the need for public services supplied by the county. (The United Way holds a large trove of data relating to demand for services provided by community organizations, as the provider of the “211” information hotline.) Since 2016, the Data Warehouse has supported DHS’s use of its Allegheny Family Screening Tool (AFST), a predictive risk modeling tool for addressing allegations of child maltreatment. The AFST uses the content of the Data Warehouse to generate a “Family Screening Score” for each call to the county’s child welfare hotline, predicting the long-term likelihood of a family’s future involvement with child welfare systems. Call center staff use the score in determining which calls to refer to investigators.

Last among these early precursors to contemporary smart city practices in Pittsburgh is real estate development. A leading Pittsburgh philanthropy funded the formation of the Green Building Alliance (GBA) in 1997 as a nonprofit organization focused exclusively on environmentally friendly building practices in the region’s commercial building sector. The GBA was the first such organization of its kind in the United States. Among the GBA’s early successes was the new David L. Lawrence Convention Center, opened in 2003, which was awarded Gold LEED certification by the US Green Building Council. That project accelerated Pittsburgh’s progress on the green building front, progress that is now linked directly to investments in smart building technology that renders the building’s energy performance more data-driven and efficient. The GBA now operates a data collection and sharing program as Pittsburgh’s “2030 District” (part of a network of “2030 Districts” around the world, a spinoff of the private Architecture 2030 advocacy organization). That program enables GBA members to collect and share data on building performance with one another and with the City of Pittsburgh.

In 2016 and again in 2019, the City of Pittsburgh added formal endorsements to these private sector efforts. The Pittsburgh Building Benchmarking Ordinance, adopted in 2016, requires owners of large nonresidential buildings to report their annual energy and water consumption to Pittsburgh. In 2019, the City of Pittsburgh adopted an ordinance that requires that all new or renovated Pittsburgh government buildings be net-zero (NZE) ready. The GBA works closely with the real estate development efforts of Pittsburgh’s Urban Redevelopment Authority (URA) and with energy planning initiatives in Pittsburgh’s commercial neighborhoods, and it partners with the CREATE Lab at CMU to develop “democratizing data” programs. Those efforts are aligned with the City of Pittsburgh’s Climate Action Plan, the first version of which was adopted in 2008. (Version 3.0 was released in 2018 following an extensive process of community engagement.) Arguably, even Pittsburgh’s legacy industrial producers are starting to get environmentally “smart” and to follow the trend toward cleaner air. In early 2021, US Steel announced that it canceled a planned $1 billion investment in emission control and production upgrades at its remaining operations in the Monongahela Valley, upriver from Downtown Pittsburgh. Instead, three batteries at the Clairton Coke Works, long the source of much of Pittsburgh’s worst particulate pollution, will be shut down. The company’s decision drew immediate and loud public recriminations from a coalition of labor unions.

Smart City Accelerants and Catalysts

Critical players and contributors to the post-2014 transition came from a variety of sources. The mayor himself stands out, though the power of his administration to move forward with smart city strategies depended in part on the fact that its interest in doing so coincided with broader national and international interest in technology- and data-based public administration, the availability of relevant technology, and political and cultural transitions in Pittsburgh.

The most important of these transitions was the new administration’s decision to create a new Department of Innovation and Performance in 2014 and a new Department of Mobility and Infrastructure in 2017. “Innovation and Performance” fulfilled a campaign pledge to modernize city administration with new technologies and practices. It is both a service center for other City of Pittsburgh departments and a coordinator of relationships with technology vendors and academic partners. Its inaugural director, Debra Lam, served in Pittsburgh until 2017, when she left to become Managing Director, Smart Cities and Inclusive Innovation at the Georgia Institute of Technology (Georgia Tech). “Mobility and Infrastructure” (DOMI) grew out of the priority assigned to modernizing Pittsburgh’s transit and transportation systems in connection with the city’s economic development goals.

Mayor Tom Murphy, who served the City of Pittsburgh between 1994 and 2006, was similarly inclined toward the uses of data and technology. But his constituency was not prepared to support a technocratic vision of Pittsburgh government, and the relevant technology was in its infancy, comparatively speaking. Murphy’s successor and Peduto’s predecessor, Luke Ravenstahl, exhibited little enthusiasm for a technology-first approach. Municipal finances compounded political and ideological barriers. In 2003 the City of Pittsburgh designated itself “distressed” under Act 47, the rough equivalent of municipal insolvency under Pennsylvania law. Tax reform and restructuring Pittsburgh’s pension system were high priorities. Pittsburgh exited Act 47 status in 2018.

Political leadership played only one role in the shift toward a more aggressive embrace of smart technologies. Pittsburgh’s smart city deployments emerged and continue to operate as complex combinations of contributions from the public sector, the philanthropic sector, local universities, private technology companies, and occasional key interventions by specific individuals. Pittsburgh’s institutional and organizational resources were summarized earlier. Additional resources partly constitute a loose network of cultural capital and partly enhance Pittsburgh’s pool of smart city expertise directly.

At the micro level, individual actors have at times played important parts in building and sustaining Pittsburgh’s contemporary technology practices. Their contributions can be traced partly to their institutional identities or affiliations and partly to their personal and professional mobility from role to role and sometimes from sector to sector, as catalysts, relationship builders, and endorsers. For example, the Allegheny County Data Warehouse was launched as part of a large reorganization of service provision by the county that included the creation of the Department of Human Services itself. The reorganization was recommended by a blue ribbon commission led by John Murray, president of Duquesne University, former dean of the law schools at both Duquesne and the University of Pittsburgh, and a widely respected community presence. The early success and longer durability of the Data Warehouse is credited both to the director of that department, Marc Cherna, and to the talent of the person later hired to expand and extend it, Erin Dalton. The success of the Western Pennsylvania Regional Data Center is partly attributable to the efforts of its director, Robert Gradeck, who helped to found and operate its predecessor organization, PNCIS, as a staff member at CMU’s Center for Economic Development. The smart cities partnership between the City of Pittsburgh and Carnegie Mellon University leaned on the experience of Richard Stafford, who directed the launch of Traffic21 in 2009 and Metro21 in 2014 and who served as the Chief Executive Officer of the ACCD from 1990 to 2003. Key individuals at Pittsburgh’s three leading philanthropic organizations have played important roles from time to time in brokering new institutional designs in Pittsburgh’s uses of public technology.

At the macro level, the City of Pittsburgh taps relationships with Results for America, a national nonprofit supporting data-based public administration and funded by Bloomberg Philanthropies; the Operational Excellence initiative and the Government Performance Lab at Harvard University, part of the Ash Center for Democratic Governance and Innovation at Harvard’s Kennedy School; and the Center for Government Excellence (GovEx) at Johns Hopkins University. Pittsburgh public administrators have been active in the Civic Analytics Network, a cohort of public data officers hosted by Harvard’s Ash Center. Metro21 at CMU spawned the MetroLab Network, a network of city–university partnerships, in 2015, as part of the White House Smart Cities Initiative.

A Smart City Catalog

Six tables of smart city initiatives in Pittsburgh follow, representing a portrait of contemporary Pittsburgh as a smart city disassembled into many of its constituent parts.

Any catalog inevitably raises classification and clustering challenges; here, those challenges are compounded by the fact that this chapter takes smart city practice to include a broad range of systems and practices. Lots of things count as smart city-related initiatives in Pittsburgh for my purposes, in the sense that lots of things are worth examining in greater detail as cases of knowledge commons governance. But they count in different respects. The classifications used below are provisional. The knowledge commons governance in evidence may be sorted differently by other researchers.

Table 6.1 lists resources and systems that constitute all or parts of smart city infrastructure. These are mostly technical systems for network connectivity and data storage, which offer the means to collect data, to combine or pool data, to access data, or some combination of the three.

Table 6.1.Infrastructure

Table 6.2 lists resources for providing citizen access to public decision-making processes, via one or more technological means. These include mobile applications for requesting public services or public information; technology platforms that provide levels of transparency with respect to public administration processes; and technology-reliant systems for soliciting community input into public decisions.

Table 6.2.Citizen access to public processes

Table 6.3 lists technology-based systems for collecting data about citizen behaviors and community resource conditions, many of which recirculate that data into decision-making processes within public administration systems.

Table 6.3.Public ICTs for citizen utility

Table 6.4 lists systems for “smart” decision-making by public authorities, consisting mostly of algorithmic processes that rely on data from a variety of sources. The line between “data governance” and “algorithmic decision-making” is fine and, in practice, possibly nonexistent.

Table 6.4.Public ICTs for data-based decision-making

Table 6.5 lists areas where technology development and deployment are parts of Pittsburgh’s public sector engagement with smart city strategies in unusual or unorthodox respects: recreation and education, on the one hand, and economic development in the private sector, on the other.

Table 6.5.Public support of ICTs in education and business

Table 6.6 lists instances of smart city practice in Pittsburgh that emanate in the first place from community engagement with community needs, in identifying problems and developing data- and technology-development strategies as governance solutions.

Table 6.6.Community data production

Does It Work?

The first and immediate question posed by any knowledge commons system and thus by a smart city governance system is whether it works. Does the system do what it is intended to do? What is it designed to do? What are its expected and unexpected costs and benefits, over different time scales? Does the system solve the problem that it is intended to solve, and does it solve a problem that needs to be solved? Does it create further problems either within its context or sector or by triggering spillover impacts elsewhere? These are not problems of knowledge commons governance or smart city technology as such. They are questions to be asked with respect to any institutional governance arrangement, and often to be asked in comparative context. How does the system work compared to one or any other actual or possible system?

Here, judgments are necessarily incomplete. Smart city practice may be motivated and influenced by ideals of effective and efficient governance, by conscious and subconscious idealization of technocratic control of urban spaces, and/or by the quest for better lives for city residents. On the ground, the question concerns the pragmatics of balancing individual and community interests, demands, and goals with available time, expertise, and material resources.

Pittsburgh’s smart city governance is flawed at least in part in the sense that some data-driven systems have been pursued or deployed without adequate consideration being given to the need to invest in complementary technologies or labor to sustain them, particularly in a highly decentralized technical configuration. Labor and expertise demands have been revealed both with respect to data analytics, statistics, and network engineering and also with respect to field-based technicians. Asking garbage collectors to carry mobile tablets to record images of potholes means reconfiguring how garbage collectors are trained and how garbage trucks are crewed. Smart street lights can’t necessarily be maintained by the same technicians who maintained older street lights.

Most of Pittsburgh’s smart city systems are too new to have been subjected to much independent review of their efficacy or effects. The Allegheny County Department of Human Services has allocated resources to producing products that assess its uses of data analytics, which are thoughtful but which are not designed to undertake comprehensive comparative institutional analysis.

Emerging descriptive research has been directed to sector-specific uses of data and algorithms in Pittsburgh governance, focusing on land use (Ghosh, Byahut, and Masilela Reference Ghosh, Byahut, Masilela and Vinod Kumar2019) and the origins of Metro21 at CMU (Preis Reference Preis2019). Some interviewees acknowledged that Pittsburgh’s use of citizen-facing technologies such as dashboards has been more complete and effective than its use of data for internal decision-making. The WPRDC is widely known in Pittsburgh and elsewhere as a model of an open data institution, but it has engaged in relatively little community outreach in Allegheny County. As noted earlier, critical approaches mostly focus on the Allegheny County AFST (Eubanks Reference Eubanks2019) and on the uses of algorithms in public decision-making in Allegheny County.

Looking at smart city governance as an element of Pittsburgh’s broader turn toward technology-driven economic development, the evidence of impact is mixed, both for better and for worse, and mostly incomplete. As with many cities, Pittsburgh often focuses on metrics that are at best imprecise, such as total dollars invested in private sector technology companies and the aggregate number of associated jobs, and at worse misleading. It is plausible to hypothesize that recent developments in Pittsburgh’s reliance on technology-based firms, and Pittsburgh’s interest in an “innovation economy,” have grown despite rather than because of coordinated or planned efforts to advance such a technocentric vision.

Experts and Expertise

Commons governance of all sorts, but especially knowledge commons, leans heavily on questions of boundaries and boundary making. Because both knowledge resources and governance resources are largely immaterial, the character of resource boundaries – including organizational and community boundaries – involves historical accident as well as institutional design, public policy choice, and logical or conceptual clarity. What Sassen refers to as borderlands are often the most interesting and important governance topic to explore (Sassen Reference Sassen2001). Few borderlands questions are as fraught, conceptually or empirically, as the question of experts and expertise in collective, community governance. The history of research science and scholarly communications, perhaps the canonical examples of knowledge commons governance through time, illustrates precisely how the role of experts and expertise has to be explored carefully in the context of broader social and community goals (Boyle Reference Boyle, Hess and Ostrom2006; Kuhn Reference Kuhn1996).

The key conceptual point, to be developed through further research, is that people working with data are almost of necessity members of functional expert communities, practicing an emergent form of knowledge commons. Community boundaries are necessarily porous; community membership is necessarily fluid. Expertise in the smart city, including Pittsburgh’s smart city, is a process of becoming, not a state of being.

The Pittsburgh smart city experience makes clear the roles of both technical expertise as to data and information technology and public policy expertise as to the uses of data in public administration and community engagement. It makes clear that those roles did not always predate the development and deployment of a range of smart city systems and strategies. Roles and their responsibilities grew and evolved over time, and the people themselves moved about the system for a variety of reasons.

Community-based expertise of this sort appears to be the hinge that distinguishes exploration of smart city practice as knowledge commons from the premise that knowledge commons governance proceeds from nuanced understanding of the role of openness in a resource management system. An emphasis on expert knowledge and expert networks, long a creature of Progressive politics (and thus dating to Pittsburgh’s earliest efforts to acquire data about itself), is in tension with that premise. As criticism of the Progressive Era makes clear, prioritizing expertise in governance of public institutions raises questions concerning democratic legitimacy that need to be parsed carefully (Hofstadter Reference Hofstadter1955). Even expert networks can be more or less open; the Pittsburgh smart city experience teaches that participating in smart city governance may require little more than volunteering some time, as in the community-based odor detection application called Smell Pittsburgh.

Pittsburgh’s expertise network is fluid enough that it is far from limited only to graduates of CMU and Pitt. But mid-level staff professionals advancing smart city initiatives in the City of Pittsburgh during the Peduto administration possess, at the least, master’s degrees.

In sum, if one of the goals of knowledge commons governance and related research is to understand how to advance overlapping goals with respect to improving the quality of knowledge resources and knowledge governance, then researchers need to carefully unpack questions of hierarchy and influence, communications patterns, legitimacy, authority, reliability and trust, accountability, and transparency. Those are all values associated with relevant expertise as such and in collective settings (Abbott Reference Abbott2001). And researchers need to pursue those questions while carefully separating them from questions of elite status or political, economic, or cultural power (Latour Reference Latour1988). In what respects are knowledge-sharing strategies imposed on the broad Pittsburgh community? In what respects is the broader community even aware of the existence of those strategies, let alone given an opportunity to voice their participation in governance strategies, by voting or otherwise contesting them? In Pittsburgh, the questions of power and elite status, and presumptive exclusion of the broad community from decisions about community welfare, were more clearly in evidence earlier in Pittsburgh’s twentieth-century experience. In the twenty-first century, the cultural authority of entitled elites has receded somewhat, but it finds echoes in the persistent influence of Pittsburgh’s largest philanthropic organizations and in the thick partnerships between Pittsburgh’s public sector and its research universities.

Hidden Intelligences

What’s missing in this account? Even a broad focus on the smart city risks missing important attributes of knowledge governance in the urban experience. In Pittsburgh, that means medicine. Undoubtedly the largest and most socially impactful contemporary data-sharing practice in Pittsburgh is not part of Pittsburgh’s smart city inventory. It is a data-sharing agreement begun in 2016 between UPMC Health Systems, the region’s largest clinical health care provider; the University of Pittsburgh, which houses a health sciences research program across six separate professional schools that is funded with close to $1 billion annually in US federal research sponsorship; and CMU, one of the world’s leading research universities with respect to computer science. This Pittsburgh Health Data Alliance, which now also partners with Amazon Web Services (AWS), feeds clinical care data from UPMC to the Alliance’s combination of medical, biomedical informatics, and computer science research communities. State-of-the-art machine-learning power is directed to developing precision medicine therapies based on nearly thirty years’ worth of clinical data. The relatively low population movement historically associated with Western Pennsylvania means that UPMC stores richer longitudinal data based on patient care than most of its peers in other US regions.

Within the medical research community, this is a highly unusual program, with extraordinary practical potential payoffs and also extraordinary ethical complexity. Outside of the medical research community, however, it is, to an even greater degree than the AFST, out of view of the broader Pittsburgh community. The only community health experience in Pittsburgh of comparably broad impact was the development and testing of the polio vaccine during the early 1950s. Thousands of Pittsburgh children accepted shots in their arms, a collaborative, public undertaking of a distinctly material and immaterial sort that Pittsburgh is proud to share and celebrate as a community triumph (Greidanus Reference Greidanus2010). The Pittsburgh Health Data Alliance operates almost entirely and solely as a function of the community of medical experts.