Dynamics in the governance of smart cities: insights from South Korean smart cities

ABSTRACT

This paper aims to explain how the governance model changes in smart cities by using South Korea's experience as a case study. South Korean smart city development is divided into three phases according to the national master plan and maturity of smart cities. Phase 1 took place from 2008 to 2013, phase 2 extended from 2014 to 2018, and phase 3 began in 2019 and will last into 2023. The cases under consideration are three representative smart cities: Seoul, Songdo, and Sejong. The analysis traces how the governance model of each city changes throughout the three phases. According to stakeholders’ roles and relationships, the governance model is categorized into four types (corporate, market, multilevel, and network governance). The governance model in all three cities showed little change. In Seoul and Songdo, market governance was dominant, while in Sejong, multilevel governance was. However, in phase 3, the institutional setting is changing to facilitate network governance in all three cities. The government guided the collaboration, giving more opportunities for private, academic, and civil actors to engage.

KEYWORDS:

• Smart city
• governance
• collaboration
• network governance
• South Korea

Highlights

• Network governance is not automatically achieved just by developing smart cities

• In the process of smart city development, governance models can change

• In the later phase of smart city development, there were efforts to engage citizens

• State-guided collaboration can be helpful when participatory governance is uncommon

• Unclear roles and responsibilities can lead to conflict and misunderstanding

1. Introduction

Smart cities aim to achieve sustainable urban development and high quality of life by applying information and communication technologies (ICT) in urban settings (Angelidou, 2015; Caragliu, Del Bo, & Nijkamp, 2009; Neirotti, De Marco, Cagliano, Mangano, & Scorrano, 2014; Sajhau, 2017; Zanella, Bui, Castellani, Vangelista, & Zorzi, 2014; Praharaj, Han, & Hawken, 2018b). Smart cities represent an integrated urban management platform, assisted by advanced technology that gathers, processes, and shares urban big data (Barns, Cosgrave, Acuto, & Mcneill, 2017; Dameri, Ricciardi, & D'Auria, 2014). The early stage of smart cities focused on sectoral ICT implementation such as transportation, security, emergency response, and utility management. For example, Adelaide in Australia and Cyberjaya in Malaysia implemented ICT in transportation and utility management (Söderström, Paasche & Klauser, 2014). More recent projects focused on comprehensive urban management, data-driven decision-making (Barns et al., 2017; Dameri et al., 2014). A smart city is a vision to achieve sustainable urban management and development.

Reoccurring characteristics of smart cities are the use of ICT and data-driven urban management and development. Yigitcanlar (2018) identified three drivers of smart cities as policy, community, and technology. Here, the policy is not limited to laws and regulations, rather governance environments that make a city smart. The governance environment actively uses technology to enhance the interaction between citizens and public agencies (Anttiroiko, 2007). Smart cities provide favorable environments to collaborate for decision-making and problem-solving by using ICT (Bakici, Almirall, & Wareham, 2013; Gil-Garcia, Zhang, & Puron-Cid, 2016). As Meijer and Bolívar (2016, p. 392) defined, smart governance is ‘new forms of human collaboration through the use of ICTs to obtain better outcomes and more open governance processes.’

In various academic fields such as urban planning, e-government, and innovation studies, smart governance has been recognized as linking urban governance and technologies to find more effective and efficient solutions to urban problems (Meijer & Bolívar, 2016). Existing smart governance literature provides insights into the characteristics (Albino, Berardi, & Dangelico, 2015; Nam & Pardo, 2011; Harrison et al., 2010; Pereira, Cunha, Lampoltshammer, Parycek, & Testa, 2017) and the role of technology in governance (Paskaleva, 2013; Odendaal, 2003; Barns et al., 2017). Dameri and Benevolo (2016) argued long-term smart strategies and a comprehensive governance framework are essential by analyzing the government structure in Italian smart cities. Kim (2015) studied conflicts and relationships among stakeholders in Korean U-city developments. These studies focus on actors’ involvement and their behavior in a cross-sectional view. However, there are few empirical studies on how smart city governance is formulated and changes along with the smart city development proceeds (Meijer & Bolívar, 2016). Praharaj, Han, and Hawken (2018a) identified four types of smart city governance: 1) traditional government promoting smart city development; 2) informed governance emphasizing smart decision-making based on urban data gathered and processed by ICT infrastructure; 3) e-governance or smart administration; and 4) collaborative smart governance which is the highest level of institutional transformation. However, as the authors pointed out, these observed governance models still pose a question about the government's role.

In that sense, this paper seeks to explain what type of governance model can be identified in the process of smart city development by taking the experiences of South Korea. The main research question is: ‘What is the governance model identified in developing smart cities?’ Smart city governance aims to facilitate collaboration among stakeholders (Albino et al., 2015; Nam & Pardo, 2011). This study serves to reveal the dynamics involved in smart city governance by analyzing institutional settings and the roles and responsibilities of related actors.

South Korea is an example that shows how smart city development has been evolved as a comprehensive urban development (Hwang, 2020; Yigitcanlar, 2015). South Korean smart city development can be grouped into three phases according to the policies: the construction phase (2003–2013), connecting phase (2014–2018), and enhancement phase (2019–2023) (Lee & Chang, 2019; Myeong, Kim, & Ahn, 2021). In the construction phase, the Ministry of Land, Infrastructure, and Transport (MoLIT) initiated ubiquitous city (U-city) developments that focused on sectoral ICT implementation. ICT is implemented in transportation, security and safety, emergency responses, and disaster prevention (Cho, 2017). As the smart city concept evolved into comprehensive urban management, the strategies also devised ‘Ubiquitous City Centers (later changed to Smart City Centers),’ providing more comprehensive smart services in public administration, welfare, education, and healthcare starting from 2014. In 2018, the government announced a new vision for smart cities, ‘Innovative Platforms,’ and appointed two cities (Sejong and Busan) as pilot cities. The early two phases follow the traditional top-down urban development process, where the central government takes charge of decision-making and execution (Hwang, 2020; Jang & Kim, 2017). This tendency is changing as the government tries to find a horizontal and innovation-driven way of instituting smart city development (Hwang, 2020).

The remainder of the paper consists of the following sections. First, the literature review introduces the characteristics of smart cities, smart city governance, and governance models. A theoretical framework is then developed to identify the role of stakeholders and governance models in smart cities. Next, the analysis uses the South Korean case to investigate the governance model in different smart city development phases. Finally, the paper concludes with a summary of findings and a discussion.

2. Literature review

2.1. Smart city and governance

Governance is a broad concept encompassing complex structures and networks in the governing process (Weiss, 2000). Rhodes (1997) defined governance as the process of ruling, cooperation, and controlling. It is ‘a social-political system as the common outcome of efforts in intervention by all actors concerned (Kooiman & van Vliet, 1993, p. 64).’ It is about balancing the public and private interests and representing citizens’ interests (Pierre, 1999). The core characteristic of governance is an autonomous collaboration by a self-organized network (Jessop, 1999; Lee, 2003; Rhodes, 1996). Governance is essential in urban planning because networks and relationships among various stakeholders are crucial to solving modern cities ‘wicked problems’ (Kooiman, 2003).

Smart cities suggest a new avenue for urban governance by using ICT to enhance collaboration and network among the stakeholders (Bakici et al., 2013; Gil-Garcia et al., 2016 (Meijer & Bolívar, 2016). The Internet and open data policy enhance access to information, encouraging citizens to engage in urban planning and public decision-making (Bisschops & Beunen, 2019). A smart city is a platform wherein the public sector, private sector, and citizens can collaborate to design innovative solutions for urban problems (Breuer, Walravens, & Ballon, 2014; Hollands, 2008). Moreover, one of the expected benefits of smart cities is facilitating good governance (Lim, Edelenbos, & Gianoli, 2019). Good governance means transparent, democratic, and inclusive decision-making processes driven by actor engagement (Afzalan, Sanchez, & Evans-Cowley, 2017; Bakici et al., 2013; Bifulco, Tregua, & Amitrano, 2017; Gil-Garcia et al., 2016; Neirotti et al., 2014).

The literature on smart city governance shows two essential characteristics. The first is an emphasis on various stakeholders’ participation (Albino et al., 2015; Nam & Pardo, 2011). Smart city governance values the consensus-building process to reach a shared agreement among stakeholders (Snow, Hakonsson, & Obel, 2016; Ruhlandt, 2018). Interaction among stakeholders can foster innovative and sustainable urban development (Nam & Pardo, 2011). In particular, citizen engagement and citizen-centric decision-making are highlighted (Fernandez-Anez, Fernández-Güell, & Giffinger, 2018). Active citizen involvement in the planning process can identify citizens’ needs and lead to the development of better services, which contribute to a better living environment. A smart city can support citizen involvement by providing a communication platform built upon the ICT infrastructure.

The second essential characteristic entails using ICT in governance to enable active participation, collaboration, and transparent decision-making (Harrison et al., 2010; Pereira et al., 2017). Smart cities use ICT for economic development, providing opportunities for change and enhancing inclusiveness (Allam & Newman, 2018). Stemming from e-governance, smart city governance aims to improve public administration efficiency, citizen-centricity, and networking between government and external stakeholders (Heeks, 2001). The information-sharing inherent in this approach allows a smooth process that provides better services to citizens, optimizes implementation, and promotes well-informed, collaborative decision-making (Soon, Luna-Reyes, & & Sandoval-Almazán, 2012). Online participatory tools encourage citizens to participate in the decision-making process (Afzalan et al., 2017; Romanelli, 2013), which contributes to improving equality (Sajhau, 2017; Yigitcanlar, 2015) and inclusiveness (M Angelidou, 2015; Gil-Garcia et al., 2016; Zygiaris, 2013).

Smart city governance emphasizes collaboration and consensus-building among stakeholders backed up by digital technologies (Albino et al., 2015; Pereira et al., 2017). To use digital technologies, it is required to implement ICT infrastructure, which involves large-scale urban (infrastructure) development. Therefore, political support and commitment are crucial elements for successful smart city development (Snow et al., 2016; Stratigea, 2012). The government plays a focal role in smart city development by providing a vision, strategies, and an open environment for stakeholder interaction and collaboration (Bakici et al., 2013; Lee & Hancock, 2012; Nam & Pardo, 2014; Rabari & Storper, 2015; White, 2016). Hence, many smart city developments take top-down approach, which seems to contradict what smart city governance pursues. For example, international ICT vendors initiate smart city projects such as Rio de Janeiro or Songdo, aiming to integrate ICT infrastructure in urban settings to provide a one-stop urban management tool. However, both cities were criticized for being a market-driven scheme and controlling everyday urban life under surveillance (Breuer et al., 2014).

2.2. Governance model

Klijn and Koppenjan (2015) identified four types of governance based on the government's role: 1) corporate governance, 2) market governance, 3) multilevel governance, and 4) network governance. Although the authors argue that network governance is genuine governance because it values a horizontal relationship as opposed to the government (public organization) taking the central role, this paper acknowledges that all types of governance are possible in the ‘process of governing (p. 6).’

Corporate governance emphasizes the government's functioning operation as public administration based on laws and rules, succeeding the traditional bureaucracy. Market governance is new public management. The government produces goals and policies and involves external organizations in executing service delivery. It uses market mechanisms such as contracts, competition, and benchmarks to increase the bureaucracy’s efficiency. Multilevel governance involves inter-governmental relations among different levels of government. Multilevel governance has emerged from the financial crisis and decentralization that made local governments less dependent on the central government (Peters & Pierre, 2011). It overcomes rigid jurisdiction problems because it enables more flexibility in problem-solving (Hooghe & Marks, 2003). Finally, network governance puts the government into complex interactions among public and non-public actors. It requires various steering strategies because the nature of the network among public and non-public organizations is complicated.

Network governance is critical in solving complex and intertwined urban problems by engaging actors with specialized knowledge and information. Network governance values autonomous collaboration through reciprocal interaction without official authority (Kooiman & van Vliet, 1993). Network governance involves ‘coordination characterized by informal social systems rather than bureaucratic structures within firms and formal contractual relationships between them’ (Jones, Hesterly, & Borgatti, 1997, p. 911). Here, the definition’s emphasis is on an informal network between the public and non-state stakeholders. In network governance, stakeholders are active and committed, not merely bounded as a consultancy.

2.3. Identifying the governance model

The governance models can be identified by the actors’ role and their interaction mode. Actors are mainly divided into two categories, governmental and non-governmental actors. Non-governmental actors include private, civil, and academic organizations (Ruhlandt, 2018). The interaction modes are categorized as participation, partnership, and collaboration based on the intensity and formality of an actor's involvement (Dameri & Benevolo, 2016; Ruhlandt, 2018). Participation has different levels according to the participant's influence on decision-making (Arnstein, 1969): informing, consulting, and placating. These three levels are in accordance with the intensity of involvement. Informing represents one-way communication from authorities to participants, notifying them of their rights, and participants’ influence is minimal. Consultation involves listening to opinions from surveys or public hearings, but the decision is still up to the authorities. Placation entails inviting a few stakeholders to the table, but authorities still wield significant power in decision-making. Higher participation rungs are excluded because participation here refers explicitly to the involvement of processes and having limited influence over decision-making.

The partnership is similar to collaboration because private and public actors work together, but its priority is not the pursuit of collective decision-making (Ansell & Gash, 2007). It is more focused on delivering the agreed-upon result than on achieving a consensus. Common examples of partnerships are public-private partnerships (PPP) (Lee, Hancock, & Hu, 2014; Odendaal, 2003) or public-private-people partnerships (Schaffers et al., 2011).

Collaboration emphasizes consensus-building and mutual decision-making (Ansell & Gash, 2007; Shergold, 2008), which smart city governance also emphasizes. The roles and responsibilities are decided upon with an open discussion among the various actors, and their relation is horizontal. Collaboration is an advanced form of interaction among interdependent but autonomous actors who share interests, responsibilities, and benefits (Shergold, 2008). Since smart city governance emphasizes all stakeholders’ active engagement in the decision-making process, collaborative interaction is desirable.

The government's roles are defined as commissioner, coproducer, and facilitator (Span, Luijkx, Schols, & Schalk, 2012). The government, as a commissioner, sets the goal and vision and takes full responsibility for the project. The government contracts out projects to non-governmental institutes (e.g. PPP projects) while setting the rules and overseeing the process. Here, the non-governmental actors participate as executors. In the opposite position, the government as a facilitator stands neutral and steers negotiation and consensus-building (Ansell & Gash, 2007). Other actors are initiators who share responsibility and power in decision-making. In the middle, the government as a coproducer shares joint goals and responsibilities with the other actors who act as partners.

Table 1 simplifies the governance models, the central role of the various actors, and their interaction modes. Considering that corporate governance comes from traditional bureaucracy, the government acts as a ruler. It provides public goods and services, and there is little room for non-governmental actors to become involved in the process. At best, their role is that of the participant, being informed or consulted. PPP is a common form of market governance (Klijn & Koppenjan, 2012). In market governance, the government commissions the project, and private actors execute it. The participation level is higher than it is in corporate governance. Multilevel governance focuses on the internal relationship of governmental agencies. The main concern is the relationship among central governments, local governments, and public agencies. In multilevel governance, the central government takes the leading role as a facilitator, while local governments are coproducers who have more autonomy than the traditional public administration. In network governance, the government takes the facilitator role, yielding room for other actors in decision-making and implementation. In this case, the actors are initiators, and their interaction mode is collaboration. As discussed in the previous section, smart city governance aims for collaboration and horizontal decision-making process that network governance emphasizes, making network governance desirable in smart cities.

Table 1. Governance model.

Governance Models	Major Role of Actors		Interaction Mode
	Government	Other Actors
Corporate	Ruler	Participant	Participation (Informing or Consulting)
Market (PPP)	Commissioner	Executer	Participation (Placating)
Multilevel	Central gov't – commissioner/ facilitator Local gov't – executor / partner	Participant/Executer	Partnership
Network	Facilitator	Initiator	Collaboration

3. Overview of smart city development in South Korea

Smart city development in Korea can be categorized into three phases (see Table 2). The first phase is represented by U-city development, a prototype of Korean smart cities, initiated from 2003 to 2013. U-cities, Korean version of smart cities, were developed by the local governments in the early 2000s, even before smart cities gained international momentum (Lee & Chang, 2019). In 2008, the Ubiquitous City Law was enacted, and many municipalities officially launched U-city projects. This period is the ‘construction phase,’ establishing ICT infrastructures for smart services (Lee & Chang, 2019). In association with telecommunication companies, the central government-initiated network infrastructure construction projects. The first phase focused on implementing smart technology in transportation, security, disaster prevention, and environment (pollution monitoring) sectors. For example, emergency response systems reduced response time by establishing a cooperation system among local police stations, hospitals, and administration (transportation) in different jurisdictions. However, U-city development was criticized for being rigid in a top-down development process that neglected social and cultural infrastructure (Yigitcanlar, 2015).

Table 2. Three Phases of Korean Smart City Development.

Phase	Phase 1 Construction Phase	Phase 2 Connecting Phase	Phase 3 Enhancement Phase
Terminology	Ubiquitous city	U-city/Smart City	Smart City
Period	200(3)8–2013	2014–2018	2019–2023
Vision	Advanced information cities that improve quality of life and urban competitiveness	Advanced creative cities that ensure safety and happiness	Smart city, an innovative platform that changes citizens’ lives
Characteristics	Focus on constructing ICT infrastructure Focus on developing metropolitan area and new cities A government-led, top-down approach	Construction of ‘Integrated U-City Management Center (Smart City Center)’ Focus on balanced development among cities Public-private partnership	Pursuit of comprehensive and innovative urban (re)development Development of citizen-centric services Encouragement of citizens’ and stakeholders’ participation
Major Urban Services	U-transportation U-security	Smart transportation Smart security Smart administration Smart healthcare Smart welfare	• Comprehensive urban services including culture & tourism, employment, education, and housing
Major Projects	U-city Testbed U-city Master Plan U-city Infrastructure Construction Innovative City Development	Integrated U-city Management Center U-city R&D U-city Profession Nurture	National pilot cities Smart City Challenge (Bottom-up solution) Smart Urban Regeneration

Source: The first, second, and third Smart City Master Plan (MoLIT)

From 2014 to 2018, the smart city projects faced the second phase, the ‘connecting phase,’ where scattered smart services are merged into one, and a new governance structure is created (Lee & Chang, 2019). Smart City Centers (the original name was Comprehensive U-city Management Center but later changed to Smart City Centers) were established to provide smart services in one go. Smart City Centers aim to provide citizen-oriented urban services, including transportation, security and surveillance, one-stop administrative services, healthcare, welfare, and education. The smart services were scattered throughout different governmental departments in the first phase, but the Smart City Centers enabled one-stop administration. However, a criticism of the second phase points out that the projects still focus on constructing infrastructure and establishing systems. Moreover, due to a lack of funding, some of the projects were stopped or postponed.

More recently, since 2019, the government has presented a new vision for smart city development. The third phase focuses on comprehending urban services in everyday life. It is yet unclear what constitutes an ‘advanced’ smart city, but the government emphasizes it is different from just implementing infrastructures. According to this notion, the smart city aims to provide a comprehensive urban experience with advanced technology. In 2018, the government designated Sejong and Busan as national pilot smart cities. They put particular focus on local needs and characteristics.

Smart cities, in theory, advocate stakeholder involvement in the decision-making process (Ruhlandt, 2018). However, empirically, South Korean smart cities are government-led economic development schemes, whereby the central government appoints several cities or districts to implement ICT infrastructure and services (Lee et al., 2014). The MoLIT initiated smart city projects and invited cities to join an open competition for financial support (Lee et al., 2014). The Ministry of Interior and Safety took charge of smart administration projects, while the Ministry of Trade, Industry, and Energy supervised smart energy and environment policy. Public agencies are also involved in smart city development (e.g. Korea Land & Housing Corporation (LH) and local public corporations such as SH). According to the Smart City Law, local governments can initiate smart city projects in cooperation with LH, MoLIT, or private sector. The law also stipulates that MoLIT should approve the master plan prepared by local governments. The procedure for making smart cities in Korea is a top-down process.

4. Methodology

The analysis focuses on 1) identifying the actors (public, private, civil, and knowledge institutes), their roles and interaction modes public, 2) identifying the governance model, and 3) tracing the changes in the governance model during the three phases of smart city development. Data are collected mainly from secondary sources, such as policy documents, laws, regulations, planning documents from governmental websites, research papers and reports, and news articles. Written articles are useful sources to trace the change over a long period. Policy or law documents reveal the government's intention, while archiving research papers and news articles provide information about the smart city development progress, including issues and conflicts.

First, overall planning information is gathered from the National Smart City (U-city) Master Plan from MoLIT and each municipality’s smart city (U-city) plans. These documents provide information on vision, goals, strategies, and specific projects. Smart City Law (U-city Law) and each municipality's ordinances are reviewed to obtain more detailed information on the process. Using this information as a baseline, information is gathered from news articles, knowledge institute's reports, and academic papers. The news articles are collected from a digital archive, ‘BIG Kinds’ (kinds.or.kr), with the keyword search ‘smart city’ or ‘U-city.’ They span the years 2008–2018. The news articles bring about contexts and relationships among stakeholders that are unwritten in policy documents. The role and interaction between government (ministries, municipalities, public agencies) and other actors (private, academic, and civil initiatives) are recorded based on the framework.

Three cities were selected for the case study (see Table 3), Seoul, Songdo (Incheon), and Sejong, which are representative Korean smart cities (Leem, Han, & Lee, 2019). In the first phase, all three cities have been appointed as U-city testbeds, and each city has gone through various smart city projects. Seoul is the most advanced city in South Korea, with a highly concentrated population and robust industry, finance, and culture. It is known for e-government and various smart city projects within the city. Songdo is a newly built city on the reclaimed land from scratch, promoted as Free Economic Zone and a ubiquitous eco-city. Both cities are promoted as successful cases of smart city development. Sejong is also a newly developed city with the purpose of balanced territory development of MoLIT, to ease the burden of Seoul by re-locating government agencies and research institutes. One of the development districts was appointed as the national smart city pilot in 2018. These cities manifest top-down smart city development. They were ‘designed and built from scratch with the optimization of urban processes through technology in mind’ (Breuer et al., 2014, p. 155). Investigating how the governance models change in these cities can provide insights into the dynamics of smart city governance.

Table 3. Case studies.

City	Songdo (Incheon)	Seoul	Sejong
Area	53.4(1,063)㎢	605㎢	465㎢
Population (as of 2019)	155,000 (2,957,026)	9,729,107	340,575
Project duration	2008–2024	2006–2020	2007–2030
Urban characteristics	Greenfield	Brownfield	Greenfield
Smart City Plan and projects	2007 Tomorrow city (U-city testbed) 2014 Incheon Ubiquitous City Master Plan(2014– 2018) 2020 Incheon Smart City Master Plan(2020–2024)	2007 U-city testbed 2010. Ordinance on U-city Construction Project 2010. U-city Seoul Master Plan 2016. Seoul Digital Master Plan 2020	2007 U-city testbed 2018 Sejong Smart City Strategic Plan
No. of news articles reviewed	51	63	58

5. Dynamics in the governance of South Korean smart cities

5.1. Seoul

5.1.1. Background

Seoul is the capital city of South Korea. It is known for its high-tech urban infrastructure and e-government services. Seoul Metropolitan Government has strong political and financial power and a separate urban development public agency, Seoul Housing & Communities Corporation (SH). In Seoul’s planning documents, different catchphrases are used, such as ‘smart urban regeneration,’ ‘smart green city,’ ‘smart energy city,’ and ‘compact smart city.’ The commonality is the incorporation of ICT in urban development projects. Seoul is also well-known for its e-government. From its various websites, Seoul provides administration services (seoul.go.kr), geographic information using 3D maps (map.seoul.go.kr), information on policy and projects (seoulsolution.kr, opengov.seoul.go.kr), and statistical data (data.seoul.go.kr). Seoul is president of the World Smart Sustainable Cities Organization (WeGO), an international agency launched in 2010 to promote e-government and smart sustainable cities for all. Currently, Seoul carries out various smart city projects in the transportation, security, environment, economy, welfare, and governance realms, mostly focusing on utilizing advanced technologies and big data.

5.1.2. Roles of Actors and Their Interaction mode

Table 4 shows the actors’ roles and the governance models of Seoul. Public actors include ministries from the central government, local governments (Seoul Metropolitan Government and municipalities), a public agency (SH), and an international agency (WeGO). Smart city development depends on the Seoul Metropolitan Government, which directly initiates various smart city projects while supporting municipalities with finance and policies. The Seoul Metropolitan Government established a smart city development base by, for example, enacting the ‘U-city construction project ordinance,’ preparing a ‘U-city (Smart City) Master Plan’ every five years, and investing public funds into projects. The Seoul Metropolitan Government was also the core of the governance model because they maintain various relationships with public and private agencies. For example, they initiated the U-Seoul Forum, a network among the Seoul Metropolitan Government (local government), KT (a private company), KAIST (university), the Ministry of Science, and ICT (central government), for U-Seoul projects in 2008. In 2016, they established cooperation with KOICA and KOTRA (public agencies) to export Seoul's e-government policy and invited IT companies to cooperate in making Digital Capital Seoul. They also hosted the Seoul International Digital Festival in 2017 in cooperation with the Seoul Digital Fund.

Table 4. Actors and their roles in Seoul Smart City.

Phases			Phase 1 (2003–2013)	Phase 2 (2014–2018)	Phase 3 (2019–2023)
Major Plans and Projects			U-Seoul Master Plan (2006–2011) Low-carbon Green Growth Master Plan Eunpyeong New Town	Smart Seoul 2015 (2011–2016) Connected City (public free Wi-Fi)	Smart City Seoul Promotion Plan 2018 Seoul Digital Master Plan 2020 (2016–2021) Smart City Seoul Platform
Focus of Development			High-speed Internet network infrastructure CCTV infrastructure City-integrated operation center	Open public information Data center Smart energy cities	CCTV operation center IoT pilot projects Citizen-centric services Sustainability and collaboration
Roles of Actors	Public	Central	Facilitator	Facilitator	Facilitator
		Seoul Metropolitan Gov’t	Commissioner	Commissioner/ Facilitator	Commissioner/ Facilitator
		Municipalities	Commissioner	Commissioner	Commissioner
		SH	Commissioner	Commissioner	Commissioner
	Private		Executor	Executor	Executor/Partner
	Civil		–	Participant	Participant
	Academic		–	Partner	Partner
Interaction Mode			Participation	Participation	Partnership
Dominant Governance Model			Market governance	Market governance	Market governance

Private actors are participating in the governance as mostly executors by contract. For example, LG CNS participated as the contractor for U-city development in Eunpyeong New Town during phase 1, and SKT provided technology solutions for Smart Seoul in phase 2 (Gang, 2010; Kim, 2017). These actors develop technology or applications and bring them to the Seoul Metropolitan Government. The final decision on how to use such technologies depends on the Seoul Metropolitan Government. It was difficult to find private or civil-initiated smart city projects.

Knowledge institutes participated with research input. Once the Seoul Metropolitan Government initiates research projects, universities or research institutes conduct research to back up the policies or projects. Direct citizen engagement has been hard to find. There are no citizen initiatives directly participating in the decision-making process. Citizens can get involved with an open competition or public hearing. For example, the Seoul Center for Creative Economy & Innovation launched the ‘Smart City Idea Open Competition’ in 2016. However, the selection and execution of this competition are still the responsibility of the Seoul Metropolitan Government. However, in phase 3, the municipalities started a public-private-citizen initiative to encourage collaboration. For example, Seoul Gwanak-gu plans to establish a smart city initiative including public and private agencies and citizens (Lee, 2018).

5.1.3. Governance model

Although the policies tried to include the private sector and citizen participation in making smart cities, the dominant governance model is market governance throughout the phases. In phase 1, Seoul Metropolitan Government took charge of the initiation of the smart city development and contracted out the ICT infrastructure construction to private companies. In phases 1 and 2, the Seoul Metropolitan Government had vast responsibilities and roles. It established the smart city project's foundation by making the master plan, enacting ordinance, and investing public funds. Putting the Seoul Metropolitan Government at the core, private agencies, public agencies, universities, research institutes, and citizen initiatives were invited to participate. It was difficult to find private or citizen-initiated smart city projects. It is mainly because of the characteristics of smart city development in Korea. Smart city development is seen as a new opportunity in new town development. Because new town development involves vast investment in construction, the cost is enormous while the return is unclear. As such, private or citizen initiatives are reluctant to invest in high-risk projects. In phase 3, according to the Smart City Seoul Promotion Plan (2018), the policy acknowledges the importance of citizen engagement and tries to establish a smart governance system. Although the Seoul Metropolitan Government still holds vast authority, it promotes citizen engagement by launching a Smart City Community Network and sharing smart city information. All in all, the governance model is market governance, a public-led and top-down approach. Still, in phase 3, it tries to encourage more participation from other actors, which shows the possibility of transition to network governance, preferable for smart governance.

5.2 Songdo

5.2.1. Background

Songdo is located close to Incheon International Airport and Seoul, a locational advantage that attracts international businesses. Songdo was initially developed as part of the Incheon Free Economic Zone in the early 2000s. However, that development stagnated due to the global economic crisis in 2008 (Shin, 2016; Shwayri, 2013). Then, the Incheon municipality decided to add the smart city concept to vitalize development. The vision for Incheon Free Economic Zone became being a ‘Global Leading Smart City.’ The government promotes Songdo smart city as a representative smart city in Korea (Leem et al., 2019). A big part of the project entails implementing ICT infrastructures and providing smart services in transportation, safety, disaster prevention, environment, and facility management. These services are provided by the Smart City Operation Center, which opened in 2014.

5.2.2. Role of Actors and their interaction mode

Table 5 shows the various actors’ roles and their interactions in Songdo smart city by phases. The central government (MoLIT) is mostly engaged as a facilitator that approves plans and provides funds. Incheon municipality established an external branch, Incheon Free Economic Zone Authority (IFEZ Authority). IFEZ Authority functions as general manager of Songdo development. Incheon municipality sets the U-city Plan (2014–2018) and Smart City Plan (2020–2024) while IFEZ Authority contracts out the projects to private companies.

Table 5. Actors and their roles in Songdo Smart City.

Phases			Phase 1 Construction Phase	Phase 2 Connecting Phase	Phase 3 Enhancement Phase
Duration			2003–2013	2014–2018	2019–2023
Major Plans			U-city plan (2003 –2009)	Incheon U-city Plan (2014–2018)	Incheon Smart City Plan (2020–2024)
Focus of Development			Developing a U-city strategy Infrastructure construction (Smart City Operation Center)	Comprehensive Smart City Operation Center (2014) MoLIT pilot project (2010–2016)	Public smart services Public-private service Pilot project
Roles of Actors	Public	Central	Facilitator	Facilitator	Facilitator
		Incheon Municipality	Facilitator	Commissioner/ Executor	Facilitator/ Commissioner
		IFEZ Authority	Coproducer	Commissioner	Commissioner
	Incheon Smart City Corporation		Initiator	Executor	Partner
	Private		Partner	Participant	Executor
	Civil		–	Participant	Participant
	Academic		–	Participant	Partner
Interaction Mode			Participation (Placating)	Participation	Partnership (with all other actors)
Dominant Governance Model			Network governance	Market governance	Market governance

The private sector played a significant role in the development of Songdo, especially in phase 1 (Kshetri, Alcantara, & Park, 2014; Shin, 2016). Two private companies, POSCO and Gale International, founded a consortium in 2002 that lasted until 2014. CISCO invested 47 million USD for infrastructure construction in 2009. In 2011, Incheon municipality and CENTIOS established a public-private consortium, Incheon Smart City Corporation. CENTIOS sold its shares to Incheon municipality in 2015, which makes Incheon municipality solely responsible for the project's execution, making dominant governance model in phase 2 and 3 market governance. These consortiums had partnerships with IFEZ Authority and developed technological solutions and strategies.

Academic initiatives participated in technical input and human resources. For example, Electronics and Telecommunications Research Institutes developed technological solutions for U-city development in 2008. Yonsei University’s Songdo Campus opened the Korean MIT Media Lab in 2011 to nurture IT specialists. The civil engagement was hard to find in the policy papers or news articles. Songdo announced a plan for developing user-centric smart services (Lee, 2018), but citizens’ direct engagement in the project was difficult to find.

5.2.3. Governance model

The significant governance model in all phases is market governance, mainly public-private partnerships. But after the private sector pulled out its shares in the joint venture, the public sector held more responsibility and risk in the project. In phase 3, as observed in the case of Seoul, the central government's smart city plan seeks more involvement from citizens’ civil initiatives. Although this cannot be directly interpreted as the governance model is network governance, at least the public agencies are setting an environment in favor of network government.

5.3. Sejong

5.3.1. Background

Sejong is a newly developed city, appointed as a multifunctional administrative capital. It was planned in 2003 when President Noh's administration announced the plan to reduce excessive concentration in Seoul and the adjacent metropolitan area. The goal was to relocate major government agencies that were located in the metropolitan area to Sejong. The relocation started in 2010 and continued while the infrastructure and buildings were constructed. The plan has three phases. The first phase (2007–2014) encompassed establihsing an urban information center and a city-wide information network infrastructure. Six essential smart services were provided, including security, transportation, and facility management. The second phase (2015–2019) was a maturing stage. Eight services were added, including over-ride signals for bus rapid transit, initiation of a big data project (2016∼2018), and identification of citizen-centric services. The third phase (2020–2030) is the finishing stage, offering citizen-centric smart services.

Like Songdo smart city, Sejong also went through many changes in planning due to political influence. Initially, Sejong was developed as a Multifunctional Administrative City. Later, when MoLIT promoted U-city, Sejong was also appointed as a U-city testbed. As U-city was changed to a smart city, Sejong's plan was also adjusted accordingly. In late 2016, MoLIT added a ‘green energy’ agenda to its planning. That’s when ‘Zero Energy Smart City’ became Sejong smart city's latest theme. In March 2018, MoLIT appointed Sejong and Busan as a National Pilot City for K-Smart City.

5.3.2. Roles of Actors and Their Interaction modes

Table 6 debriefs the actors and their roles in Sejong smart city. The public sector played a focal role in the development of Sejong smart city. MoLIT designated Sejong as a smart city while LH commissioned U-city testbed project to the private sector. The National Agency for Administrative City Construction (NAACC), a public agency, was founded in 2006 to execute that plan. NAACC independently operated the construction of Sejong and was also responsible for administration in Sejong until Sejong municipality was officially initiated in 2012. After that, NAACC became a department of MoLIT but remained in charge of the urban development of Sejong. Because of this dual authority, Sejong municipality and NAACC conflicted over the project's leadership. In December 2016, they announced the same project plan as if each institute would initiate and execute the plan. The unclearly defined roles and responsibilities brought about this conflict. Sejong municipality said they announced the smart city plan to share information with citizens. The reason for not consulting with NAACC beforehand was to save time to respond to citizens’ questions. However, since urban planning is the responsibility of NAACC, NAACC reacted to the municipality as intruding on their plan. Later, in 2018, NAACC, Sejong municipality, and LH agreed to collaborate for 1) planning Sejong smart city, 2) improving regulations, 3) gathering ideas from the private sector, 4) building new governance that engages citizens, and 5) sharing data for data-based smart city operation.

Table 6. Actors and Their roles in Sejong Smart City.

Phases			Phase 1 (2003–2013)	Phase 2 (2014–2018)	Phase 3 (2019–2023)
Major Plans			–	Sejong Master Plan 2030
Focus of Development			Establish urban information center Establish basic smart services	Addition of more smart services Big Data	Citizen-centric smart services Zero Energy Smart City
Roles of Actors	Public	Central	Facilitator	Facilitator	Facilitator
		Sejong Municipality*	Partner	Partner	Partner
		NAACC	Partner	Partner	Partner
		LH	Commissioner	Partner	Partner
	Private		Executer	Executor	Participant
	Civil		–	Participant	Participant
	Academic		Partner	Participant	Participant
Interaction Mode			Participation	Participation/Partnership	Partnership
Dominant Governance Model			Market governance	Multilevel governance	Multilevel governance

*Sejong municipality was officially launched in 2012.

The other actors’ role is limited to that of the participant in phases 1 and 2. The private sector is also involved in the project but mainly as an executor by contract. For example, SK Telecom Consortium executed U-city development under the supervision of LH in 2008 (Huh, 2018). They later, in 2016, participated in developing a security system for socially disadvantaged groups as part of the smart city development (Hong, 2016).

In phase 1, there was consortium between government-funded knowledge institutes and private sectors to launch U-city testbed in Sejong (Jung, 2009). Research institutes such as the Korea Research Institute for Human Settlements provided policy recommendations and suggestions for planning strategies in phase 2 and 3. Civil initiatives did not engage in the decision-making process. NAACC launched the Smart City Living Lab, where the private sector, academia, and citizens can participate in 2018 (Oh, 2018). Also, Sejong smart city initiative was launched, composed of eight academics, five public agents, seven private agents, and 75 citizens (Hong, 2018). Members of this initiative discussed seven strategies of the Sejong smart city master plan, including mobility, health care, education, energy and environment, and governance.

5.3.3. Governance model

The dominant governance model in phase 1 is market governance. The public agencies such as LH and NAACC initiated the project and contracted implementation of U-city testbed to private-knowledge institutes consortium. In phase 2 and 3, multilevel governance and market governance are observed. Different public agencies (MoLIT, Sejong Municipality, and NAACC) are involved in the development process because Sejong is a politically appointed new city. Like Songdo and Seoul, Sejong also tries to involve the private sector and communities in the planning process in phase 3.

5.4. Findings

The first finding is that all three cities showed little change in their governance model as smart city development progressed. However, their planning documents emphasized citizen-centric services and people’s involvement. Seoul initiated the projects and contracted out the construction to private companies. Incheon Municipality and IFEZ Authority had a strong partnership with private actors (network governance), but private companies reduced their financial contribution due to economic difficulties. After the private sector revoked its investment, the public actors became the main agent in the second phase (market governance).

Sejong's governance model is a mixture of multilevel governance and market governance (PPP). Different government departments were involved in planning while the project's execution was contracted out to private firms.

In the third phase, all cities tried to engage other actors in the decision-making process, acknowledging the importance of collaboration. The public agencies initiated public-private-citizen initiatives (e.g. Living Lab in Sejong). In general, smart city projects take a long time (more than ten years in the Korean case), accompanied by political, societal, and economic changes. Also, the smart city concept has evolved into comprehensive urban management and development that accounts for all actors’ collaboration. These factors influence the projects in terms of their vision, goals, strategies, and governance.

The second finding is that all three cities acknowledge the importance of citizen participation in the later phase of smart city development, as the central government’s plan recognizes the importance of citizen involvement. This sums up that Korean smart city development is still a top-down development. Therein, public agencies took a significant role in decision-making while other actors were participants at best. Citizens’ role was limited to participation in public contests or events that public agencies organized. However, smart cities pursue citizen engagement and collaborative decision-making (Fernandez-Anez et al., 2018). As the smart city projects progressed, all three cities started to find ways to encourage citizen engagement. Still, citizen engagement is promoted by public actors, which is far from an autonomous collaboration in network governance. However limited, it is a meaningful development that citizens can become involved in smart city development based on the government’s guidance.

The third finding is that unclear roles and responsibilities can lead to conflict and misunderstanding. For example, in the Sejong smart city project, Sejong municipality and NAACC conflicted over the project’s ownership. NAACC is in charge of smart city and urban development projects aimed at making Sejong a National Administrative City, but Sejong municipality is also involved in smart city development in terms of policy and strategies. The ambiguous hierarchy in local government and public agencies has prompted this conflict. NAACC was in charge of Sejong city's administration before Sejong municipality was officially launched in 2012. The administrative responsibilities have since transferred to Sejong municipality, but urban planning and development remain the responsibility of NAACC. This overlap has made the hierarchy between NAACC and Sejong municipality ambiguous since Sejong municipality also has a department for urban planning and development. On the other hand, in Songdo, Incheon municipality and IFEZ Authority have clear boundaries on their roles and responsibilities. IFEZ Authority is an external branch of Incheon municipality that is solely responsible for Songdo smart city development. Incheon municipality also has an urban planning department but leaves the development of Songdo to IFEZ Authority. This case study shows that responsibility needs to be drawn to prevent overlap and conflict.

6. Discussion

From a governance perspective, Korean smart cities’ development follows the rule of government rather than governance (Kim, 2015). Smart cities quickly get hyped because they seem to be an attractive and visible plan to governments and private companies (Hwang, 2020). However, initial smart cities (U-cities) were seen as construction projects rather than holistic urban planning. That is because the development focused on ICT infrastructure, which is also the case in all three case-study cities. With ICT infrastructure, smart cities in the first and second phases provided sectoral smart services in transportation and safety. Still, citizens hardly noticed they were living in a smart city because those ‘smart services’ were a technological upgrade from existing services (Hwang, 2020).

Although this tendency laid the technological foundation for smart governance, governance itself was still traditional and hierarchical due to the longstanding state-initiated planning tradition of South Korea (Shin, 2016). In the first and second phases, the public sector led the project and decision-making while other actors contributed as participants. For example, in the first phase, the central government suggested 228 U-services to ease the project, but local governments pick U-services from a list rather than seeking innovative localized solutions (Hwang, 2020). In the third phase, the government introduced a new vision, ‘smart city as a platform,’ a holistic approach that engages private and civil actors’ active participation (Hwang, 2020). The emphasis moved to software (governance and network), which reflects smart cities are not only about technology but also about human and institutional capital (Hollands, 2008; Nam & Pardo, 2011; Neirotti et al., 2014). In all three cities, the municipalities tried to engage the private sector and citizens more in the development. Each city is associated with the private sector in developing technological solutions that are appropriate to the city while consulting citizens on developing citizen-centric smart services.

Although the case studies showed little change in the governance model, governance in the third phase can be seen as a stepping stone or preparation to network governance. Network governance is a horizontal network where all actors share responsibilities and active roles, and the network is autonomously formed (Klijn & Koppenjan, 2015). In South Korean cases, collaboration is promoted by the government. The public agencies provide a platform that encourages networks among private, academic, and civil actors. State-guided collaboration can be useful when planning culture is not used for participatory planning, and network is not formulated autonomously. In state-guided collaboration, the government acts as a facilitator but holds responsibility for carrying out the project to the end. The government promotes communication among the actors and reflects their opinions in the planning process. This approach can avoid responsibility issues in network governance and also, state provided overarching goal and vision can support smart city development in the city level (Praharaj et al., 2018b).

7. Conclusion

Smart cities support people in being more informed, engaged, and connected (Angelidou, 2015). They pursue collaborative decision-making using ICT, emphasizing citizen involvement and a participatory approach (Angelidou, 2015; Caragliu et al., 2009). ICT enables mutual communication among stakeholders (Gil-Garcia et al., 2016; Schaffers et al., 2012), and it provides open environments for information sharing that promote transparent and democratic processes (Angelidou, 2017; Wiig, 2015). However, as seen in all three Korean smart cities, the state still has substantial authority in smart city development. As the smart city notion evolves to be more than a technological solution, the planning also tries to facilitate collaborative governance, but it was hard to achieve directly in practice. Instead, the institutional settings are open up to citizen engagement and the private sector’s participation.

This paper has some limitations. First, the data source relies on secondary data, policy and law documents, research papers, and news articles. Policy or law documents already represent the government's agenda, which may not reflect other actors’ involvement. However, this limitation was counterbalanced by using research papers and news articles that report processes, stakeholders’ behaviors, and conflicts. Second, the three cities are well-known success cases. Other smart cities are relatively unhighlighted because their process is either delayed or not different from non-smart cities. Researching why these cities were not as promoted as Seoul, Songdo, or Sejong can reveal what makes a smart city smart and how smart governance is realized.

Despite these limitations, this paper draws meaningful insight into governance in smart cities. First, it contributes to identifying the governance model of smart city development based on the actors’ roles and interaction modes. It showed that network governance that smart city pursues is not directly achieved by developing smart cities. Smart governance needs governance innovation, a systematic change in the government in urban planning, and public services procurement. Smart cities are collaborative communities that bring public, private, academic, and civil actors together (Snow et al., 2016). Second, this paper suggests that state-guided collaboration can be a transitional model to network governance when autonomous collaboration is not well established. Although the smart city governance in Korean cases still showed top-down mechanism, the public agencies tried to engage citizens and other actors with Living Lab and initiatives. The government can take the initiative to create an open environment for other actors to engage in the decision-making process. This can also be advantageous in avoiding the responsibility issue, a weakness of network governance (Benner, Reinicke, & Witte, 2004). The government can influence the development path by initiating an appropriate vision and strategies that balance citizens’ needs and development incentives.

This paper's contribution can be further extended with research on the governance of smart cities in other countries. Different planning cultures can have various governance aspects. Also, more attention is needed on how governance models influence the development of smart cities. Which governance model is useful in making smart cities under what circumstances? Future research can also focus on the relationship between the governance model and the performance of smart cities.

Disclosure statement

No potential conflict of interest was reported by the author(s).