Large-scale land development of landed housing and apartment based on developers networking in Jabodetabek

The immense development of residential land in Jabodetabek and unreasonable property prices indicate an uncompetitive housing market. It argues that the phenomenon of oligopolistic market condition, either in the property or land market, is bridled by a minority of developers. It is reinforced by the emergence of a network formed among the development firms. In order to deliberate developer’s network, we incorporated institutional land development study approach, market structure and business competition theory. Subsequently, social networks analysis among large-scale housing developers will be explained. This study revealed out that the development of residential land was found to lead into a more competitive climate with monopolistic competitive market type. Four investigations: multiple positions; stock ownership; consortium; and kinship, were selected to divulge a large-scale housing developer network that are formed either directly or indirectly. The network created a semi-oligopolistic type of land domination. A few well-known developers became the dominant actors through landed housing and apartment developers network. The emergence of this network phenomenon is indicated by poor regulations and its institutions. Moreover, policies regarding spatial planning and business competition have not been integrated yet.


Introduction
There is no doubt that the rapid pace of land development issue emerging in Jabodetabek always rouses an interesting discussion.Prior to the reformation period, residential land and property market in Jabotabek was distorted by informality phenomenon which occurred in almost every aspect of land development process [1][2].A large majority of land acquisition over 1,000 hectares were ruled by only 15 developers.It was obvious to make a presumption that the housing market was supply-driven.A study also revealed that around 16.6 thousand hectares of rural area in Jabotabek had been transformed into residential use in only 20 years [1][2].In between 1974 and 1989, more than a half of Jakarta's land housing development permits was owned by just 16 developers [1][2].This fact is supported by a study that examines that the land and property market in Jabodetabek is controlled by a few large-scale housing developers, triggering an oligopolistic land possession structure [1].Moreover, a network has been formed between these developers.The property market was very distorted by the situation of developers in Jabotabek who cooperated in the ownership of stock shares, interconnected through kinship, etc. Cooperation between these developers tends to get stronger for housing developers with large asset capitalization scales [4].On the other hand, there is only a handful of research that investigate the behaviour of housing companies in Indonesia, most of which particularly reviews how the housing developers are linked eto others.Academically, this research is expected to provide an overview of the land development in actorbased housing in the post-reform period.The phenomenon that can be observed is a description of the network of large-scale housing developers in Jabodetabek and their land possession structures.With this study's relation to the developer network, we hope for its practical use as a consideration for stakeholders in evaluating the climate of property business competition and its hook-up to the urban planning domain.Possessing reliable mapping data will make it easier for policymakers to create more equitable policy briefs for the public in terms of land and property rights.We also observe the condition of the developer's current land possession so the result can be used to optimize efforts in providing public residence.In addition, the housing backlog in Jabodetabek over 2019 reached up to 2,716,881 units or around 30% of the total households in Jabodetabek [5][6][7][8][9] [10] [11][12] [13].
Given the above context, this paper will deliberate the magnitude of large-scale housing development conditions recently.The institutional approach of the land development process is conducted to explore the characteristics of connection among companies involved in residential development.In the next part, it will review land market structure which is segmented into each developer groups.However, the main discussion in this paper is about the identification of inter-developer networking.This section will unveil the characteristics of the company network linked through four main attributes.In the last part, this paper will provide several propositions to assist the government to deal with these challenges.

Institutional-based of land development
Past studies about land development utilize a neo-classical approach.This view points out the free market system as the determinant for any kind of economic activity including land development process [14].Land allocation in a city is assumed as a commodity that can be allocated through price competition towards land usage.Market power, including supply and demand balance, apparently does not produce much effect.There are some market assumptions that this system works without the role of any actor, particularly the government [15].The result of every interaction taking place in this system would economically determine the pattern of land allocation in a city.Notwithstanding, the simplification of this theory is exceedingly widespread; indeed this paradigm leans to dominate the derivative models of urban land allocation [16].
In stark contrast, however, there is a new theoretical approach incorporating institutional role into land development determination.This new model covers the complexity of processes and events, structures in decision-making, and actor variables with all their roles, relationships, and interests [17] [18].It is known that institutions have an important role in land development practices [19].The practice involves a lot of basic actors [20].Many cases of actual development are related to many aspects, including politics, economics, and even socio-cultural intervention that affects the decisionmaking process [21].In particular, this view is concerned with how actors could manipulate the free market mechanism [22].Additionally, the institutional approach is claimed to explore thoroughly processes and phases of land and residential development contemporarily [17].
A recent study examines, the goal of housing developer companies is to generate the maximum possible profit from land development activities [23].Therefore, the study of networks between developers can also be approached by the theory of inter-company networks [24].Referring to the pattern of connectedness between housing development companies, social network analysis is used which studies the description of linkages, connections, and relationships between actor entities [25] [26].

Land market structure
Land possession by developers shifts the market structure from monopolistic into perfect competition [27].This market structure classification is applied to identify the situation of the developer's land possession, ranging from monopoly to loose oligopoly.In the review of the law on the Prohibition of Monopolistic Practices and Unfair Business Competition as enacted in Law No. 5/1999, it is stated that a monopolistic situation occurs when a sole company or group of firms rules more than 50% of the market.
Meanwhile, in oligopolistic market conditions, there are classifications elaborated more specifically by dividing oligopoly into three levels; dominant company, strict oligopoly, and loose oligopoly.In a dominant company oligopoly, one company controls more than 40% of the market and does not have close competitors.In a strict oligopoly, there are a small number of companies that control up to 60% of the supply in the market structure.
A strict oligopoly occurs when there are eight or more companies that control up to 50% of the resource supply, and the largest company controls 20% of the supply or more [28] [29].The dominant firm market model and loose oligopoly tend to create the same price patterns as a monopolistic market, owing to the domination of firms that holds the power to control the market (Sterling et al., 2006).Another classification is a loose oligopoly, defined as a situation where 20 companies or less control up to 75% of the supply in a market, but no company supplies more than 10%-15% [28] [29].A loose oligopoly is a relatively competitive market form in the structure of monopolistic competition [30].
On the other hand, the classification of land possession based on networks between actors determined through the classification of the Sherman Act Antitrust statute (1890) ranging from oligopolistic to nonoligopolistic network types [29].A network relation between developers which is suspected to be oligopolistic betides when a firm's network has controlled over the land supply of more than 75% of the total area.If land possession in this network is only between 50% -75%, it can be argued to be nonconclusive, meaning that it cannot be concluded that the condition is oligopolistic or monopolistic.In this case, it can also refer to the semi-oligopolistic network type.Meanwhile, if the land ownership is less than 50%, it can be concluded that the network among developers is non-oligopolistic.

Inter-developers networks
A social network is a form of the underlying link of relationships between one entity and another, both at the individual and organizational level [31].Relationships between these social entities create communication channels and interactions that link between actors.This phenomenon shows an exchange of information flows, knowledge, and resources owned by each actor [32].The relationship formed between actors can create an alliance structure resulting in innovation [33].Additionally, competition between companies in a competitive environment allows the improvement of access to resources, reduction of asymmetric information, an increase of bargaining level against market pressure, lobbying against the government, and capabilities of a company as well [34].
The phenomenon of dependency between companies in terms of resources compels the creation of a network between companies.Companies always need help from others, since not all developers have qualified resources, so that the network relationship will counterbalance each other [35].The existence of this interdependence reduces the operating costs of the companies, thereby increasing production efficiency.A company's dependence on a resource can be reduced if it can establish itself in its environment through networking and cooperation [36].This environment is in the form of other actors in the land development process, including housing developers.According to this opinion, a company will be able to survive or even become more successful if it can control existing resources [37].
The existence of an increasingly competitive market certainly results in the development of a company's economies of scale.Competitive market conditions can create greater uncertainty in a company [38].Therefore, to create a sustainable strategy for the sustainability of the company in the future, various ways and efforts are carried out by the company, one of which is by controlling or intervening in other companies, so that uncertainty can be minimized [39].Therefore, this study was only used four types of relationship as the basis of determining the social networks that formed between actors, as follows : 1.Multiple Positions; 2. Stock Ownership; 3. Consortium; and 4. Kinship Relations.
Multiple positions, is a condition where a person who holds a certain position or power as a director/commissioner in a company concurrently holds another position as a director/commissioner of another company (Law No. 5/1999) [40].The consequence of this relationship is that it could facilitate collusion, that is, facilitating communication between competitors.This could also serve as a co-optation medium in strengthening their respective companies.Stock ownership as an indication of a relationship between companies, because with ownership of shares, a company can take control of decisions and access to the network in said company, and is also stipulated in Law no.5/1999 [40].The existence of share ownership in different companies in the same market allows for more homogeneous decisionmaking and may distort the market.The bidding price may be regulated unilaterally.
Consortium deliberated how two or more companies create a legal organization or project with joint ownership tied to achieving a specific goal [41].This co-ownership can also align the business decisions of the corporate alliance involved.While kinship revealed the existence of familial ties in cooperation between companies can minimize operational and transaction costs to the point of null.In addition, the process of exchanging resources is considered to be more effective because of the similarities in knowledge backgrounds with one [42].An example of this phenomenon is shown in the practice of guanxi business networks, a concept of cooperation through personal connection by Chinese society [43].
In a market with the same commodity, companies will tend to compete with each other to get the maximum profit possible.Regarding the interactions and relationships that occur between actors in the market, it will allow the exchange of important and sensitive information, especially regarding business [44].The exchange of information can strengthen the company's external stability in coordination.This external stability can provide responses and interventions related to new players entering the same market, giving rise to anti-competitive business practices [45].

Social network analysis
The prevalence of relationships emerging between actors involved in a market with the same commodity can catalyse the formation of a social network.Any relational affinity could possibly result in such connection, such as kinship, friendship, association, affiliation, short-term cooperation, alma mater, and others.Simple modelling in describing the form of a social network aims to conjugate each actor identified as having ties to another through certain indicators [46].
The principles used in analysing developer networks in this study were adapted from Haythornthwaite (1996) using at least four principles [47] as follows.1. Prominence: This principle describes which actor has the central role, which is divided into two parameters, namely: a. Degree Centrality: The centrality of actors based on the number of direct relationships indicates access and control over information and resources in the network.b.Closeness Centrality: The centrality of an actor's position on a network, signifying the rate at which the actor spreads information.2. Range: This principle indicates how broad an actor's network is and is divided into two measurements as follows: a. Reachability: The relationship of an actor with others, indicating the actor's ease of access to the social capital in the network.b.Geodesic Distance: Indicates how close the actor is to the social capital in the network.3. Brokerage: This principle indicates how an actor connects lines to other actors both on a network and on another network.Furthermore, it is divided into two parameters: a. Betweenness Centrality: Indicates the size of the actor's control of interaction and flow of information.b.Structural Hole: Indicates various alternative sources of information about actors.4. Cohesion: This principle describes the closeness of network between the actors, divided into two indicators: a. Density: Indicates the size of opportunity for interaction in the network.b.Clique: Indicates the cohesiveness of an actor group consisting of at least 3 actors that function as backbone in the network.

Methodology
This descriptive study is conducted through quantitative and qualitative approach.Phenomenon observed within this scope include: developer profiles; land distribution among developers; connections between developers; and the characteristic of formed network.Quantitative approach was used to analyze the magnitude of land acquisition of each developer groups.Furthermore, we appraised the main parameters in order to show the networking characteristic.A qualitative approach was used to unearth and analyse any kind of information mainly from secondary resources.These resources consist of multiple data compilation and inferred into the diverse ties between developer actors instead of concluding information from interviews with representative actors.The data was largely gathered through web scraping method from housing market platform, newest REI (Real Estate Indonesia) directory database, Indonesia Stock Exchange (IDX) summaries, property news and articles, journal, property magazine and tabloid, and additionally other resources considered to be relevant.Scraping method was applied to rumah.com, which has the largest property marketplace in Indonesia.The number of housing projects, name of their developers and offering price of apartments were collected from this method.On the other hand, non-random sampling was used to find respondents who were then interviewed by purposive order.The selection was considered from the precedent of market structure power in the developer's land which had been studied by Kartiwa (2015) [48].Based on this sample, there are three representative of developers that had been interviewed : PT Bumi Serpong Damai (BSD), PT Agung Sedayu, and PT Alam Sutera.
There are three objectives in this research including three main analysis.First and foremost, the exploration into any literature and information resources were undertaken through content analysis method.Afterwards, spatial analysis was conducted to explore existing housing projects area in property marketplace and some of land banks that can be mapped.Several cases utilized toolboxes in ArcGIS and MyMaps platform to measure the comparison of land acquisition between developers.Furthermore, the data and information that had been gathered were analysed by social network analysis (SNA) to reveal the relationship between actors in some similarities.The ties among developers in this study were limited into four types : multiple positions; stock ownership; consortium; and kinship.This analysis was supported by UCINET as network builder and SocNetV as visualizer.
Identification of large-scale housing was distinguished by two residential types, landed housing and apartment type.In the landed housing type, we limited the exploration of data to residential projects with 100 hectares of minimum developed area.The standard was adapted from a study in Vietnam [49].Until 2012, developers with more than 10 years of experience in Vietnam had land possession for housing of at least 100 Ha which was later referred to as a type of large-scale housing developer.Meanwhile, large-scale apartment was plotted by the highest value of offering price of at least 25 billion rupiahs and was determined by the top quartile of web scraping result.

Large scale housing projects distribution
Landed housing-type land development in Jabodetabek takes the total of 1,161 projects with expanded area of 93,520.08hectares.Landed housing projects were developed by 158 developer groups and 492 non-grouped.These large-scale landed housing (at least 100 hectares) consist of 123 projects developed by 86 companies.Large-scale landed housing development in Jabodetabek with a minimum project area of 100 hectares (10.6% of total site housing projects in Jabodetabek) controls up to 69,891 hectares out of 93,520.08hectares of expanded area, or 74.73% of total housing project area in Jabodetabek.The largest distribution of large-scale landed housing projects is in the Bekasi area, with a contribution of 41 projects.These projects occupy 26,608 hectares or 17.93% of the total area.The mean percentage of project distribution area is located in DKI Jakarta.Meanwhile, the fewest of large-scale projects are in Depok area with only 5 projects covering an area of 1,305 hectares (6.52% of the total area).Comparing all formal landed housing projects in Jabodetabek, the total development accounted is 15.51% of the Jabodetabek area.The number of project distribution in each region is quite even with around 200 projects.DKI Jakarta has the highest percentage of residential development areas with 22.77% of the total, while Bogor Region has the lowest contribution of only 7.4% of the total.However, by comparing large-scale projects to all housing projects in Jabodetabek, it can be concluded that 74.8% of the land area for housing projects is large-scale type.
Meanwhile, large-scale apartment development in Jabodetabek with a minimum unit price of 25 billion rupiahs sprawled into 45 projects and were developed by 38 firms.The growth is concentrated in South Jakarta with 25 projects.The highest unit price that can be detected is 60 billion rupiahs, attributed to Anandamaya Residence which was developed by PT Brahmayasa Bahtera as a subsidiary company from Astra Group in Central Jakarta.Almost all large-scale apartment projects are located in DKI Jakarta with a total of 41 projects.The distribution in DKI Jakarta has the following details: 3 projects in West Jakarta, 7 projects in Central Jakarta, 1 project in North Jakarta, and South Jakarta takes the most by 30 projects.In addition, 4 projects outside DKI Jakarta are located in South Tangerang City with 2 projects, and 1 project each for Bekasi City and Bekasi Regency.Therefore, the distribution of apartments worth 25-40 billion rupiahs is spread over 23 projects, 41-50 billion rupiahs totaling 15 projects, and apartments worth over 50 billion are consisted of 7 projects.Large-scale landed housing development which only contributes 10.6% of the total landed housing projects with a minimum project area of 100 hectares occupies 69,891 hectares of the total 93,520.08hectares (74.73% of land area for formal housing projects in Jabodetabek).The total land ownership for housing projects by grouped developers is 80,480.52ha or about 85.63% of the total project area.Meanwhile, developers without group control 13,039.56ha or around 14.37% of the project area.Sinarmas Group holds the largest land area of 14,067 hectares or around 15.04% of the total area of the Jabodetabek housing project, followed by Ciputra Group in 2 nd place with 8.99%, and in 3 rd place is Damai Putra Group with a percentage of 6.43%.

Inter-developers network
Large-scale housing project developers identified were then traced to each other based on 4 predefined relationship criteria.We define direct relationship when a condition fulfills the criteria for relationships that occur between large-scale housing developers in Jabodetabek without going through a liaison actor.Meanwhile, indirect relationship results from links formed with by intermediary actors/third parties.Search results from various sources are as follows: A number of 86 development companies that developed 123 large-scale landed housing projects were further analysed to identify the relationships that were produced and the network of actors that was 10 formed.It was found from the trawl that 59 developers are connected directly and 43 developers are connected indirectly.There are 4 separate subnets on the network with direct connection as well as 2 separate subnets on the network with indirect connection.Besides, among 38 large-scale apartment project development companies in Jabodetabek, 34 developers were found to be connected to each other in the context of a direct relationship by creating 1 large sub-network.Meanwhile, a total of 26 developers are connected indirectly with 8 separate subnets.In the landed housing development network with indirect links, 49 intermediary companies were identified, widening development network and complicating the detection of existing connection.In a large-scale residential development network based on indirect relationships, a total of 92 companies were formed, 49 of which are intermediary companies connecting 43 large-scale landed housing companies in Jabodetabek.Meanwhile, in a largescale apartment developer network based on indirect relationships, there are 46 connected companies with 20 intermediary companies found.
The relationship between one developer and another may be linked through more than one subnet.A subnet is a group of connected developers.The following visualization provides an overview of the subnet classification and the names of the developers.In addition, the number of relationships between an actor and other actors are also shown.The network of developers with indirect relationships is tangled with numerous intermediary actors, allowing developers to connect through a chain that is more difficult to be identified.

Land domination of developers
Information regarding the area of land possession for landed housing projects then evaluated through an assessment of the land market structure to determine the market condition in Jabodetabek using this diagram.

Figure 7. Diagram of Housing Land Possession Structure Assessment [27][29][30][40]
From previous calculation data, we can describe the characteristics of the land possession structure of large-scale landed housing developers in Jabodetabek.In the assessment of the monopolistic and oligopolistic market, dominant companies do not meet the requirements because none of the developer groups control either 40% or 50% of the total land area.Therefore, we needed to make further assessment to determine whether the market is a strict or loose oligopoly.Between the eight largest development groups, a land area of 50,454 hectares or 53.95% of the total area of housing projects in Jabodetabek is shared.This means that the characteristics of strict oligopoly market are also not fulfilled because despite eight or less developers/development groups control more than 50% of the land area, the largest developer, the Sinarmas Group, does not control at minimum 20% of the land area.The next assessment was to check whether the land possession structure is a loose market oligopoly.We found that 20 developer groups with the largest land hold controlled 67,626.2hectares of housing project area in Jabodetabek, approximately 72.31%, just barely reaching the 75% threshold.Although one developer group, namely the Sinarmas Group, reaches land ownership percentage of 15.04%, which is 0.4% or about 37.4 Ha larger than the appraisal requirement of 15%, this still leaves the conditions for a loose oligopoly unfulfilled.From the results of this assessment, the conditions met in the landed housing market in Jabodetabek is of the relatively competitive land possession market.In accordance with the continuum of market characteristics, this condition will increasingly leads to a monopolistic competition market with relatively more competitive companies and a market that is easy to enter [30].This market structure also allows at least 500 identified housing businesses in Jabodetabek to determine the price for each housing product with relative ease.Furthermore, we assessed the characteristics of the network arising between developers.The evaluation of characteristics in this network was adapted from the Sherman Act Antitrust statute ( 1890  Identification of land possession in the developer's network was carried out in each subnet as a representation.Based on the four subnets formed in the developer network with direct links, requirements were met for semi-oligopolistic network.This can be seen by observing land ownership, where the largest sub-network contributes 52.69% to the total land area for housing projects.This percentage is still higher than the requirement for the lowest semi-oligopolistic network of 50%.Therefore, the characteristics that are met in the network between developers with direct relationships is of a semi-oligopolistic network.Meanwhile, the identification of network characteristics among developers with indirect relationship found networks with land possession above 50%.The largest land ownership contributes 50.54% to the total area of housing projects in Jabodetabek.This indicates that the characteristics of the network formed through indirect relationship is also semi-oligopolistic.

Policy related to developers network
The results of this study indicates that the land possession of every developer has led to a more competitive direction.However, it is also shown that network between development actors was still formed, raising the presumption of an anti-competitive market.Law No. 5/1999 has prohibited various forms of relationship such as concurrent positions and dominant share ownership which are deemed to potentially be able to distort the market [40].The regulation also mandates the types of agreement, activities and dominant positions that are prohibited.The implementation of this law enforcement is carried out by a commission called the Komisi Pengawas Persaingan Usaha (KPPU).This commission has the authority to investigating, prosecuting, consulting, examining, witnessing and deciding cases.
IOP Publishing doi:10.1088/1755-1315/1318/1/01201614 However, KPPU does not have the authority to prosecute and impose laws on perpetrators of unfair business violations.The matter of imposing both criminal and civil sanctions is under the authority of the Pengadilan Negeri (District Court) and the Pengadilan Tinggi (Supreme Court).KPPU's position is more of an administrative institution, so that sanctions given are also administrative in nature.
In reality, KPPU has several main obstacles in monitoring business competition practices [50].Firstly, KPPU does not have the authority to conduct search into companies suspected of having unfair business practices.Secondly, the confidential nature of company data makes it so that the KPPU is unable to carry out examinations and investigations optimally.Third, KPPU has not been able to coordinate well with other government agencies.Lastly, KPPU cannot compel the presence of witnesses and business actors, even though they have the authority to summon them.
In practice, KPPU is often shown to be suspicious of companies with large capitalization.However, investigations carried out on business actors at large companies often reveal information gaps.This is due to the large amount of company documents and portfolios that need to be processed, and there is always the possibility of documents that are kept confidential.Unfair law enforcement practices are more commonly found in cases where the legal prohibition is written (per se illegal).However, investigations carried out on cases based on the consequences of a business practice (rule of reason) is found to be very rare [51].Meanwhile, the existence of a network between development actors that raises the presumption of an unhealthy market cannot be investigated tangibly.Network impacts can be covert, deeply-rooted, and long-term.This of course often escapes the legal perspective.
Other relationships that may form networks between developers, such as consortium practices and kinship between companies, are not stipulated in Law No.5/1999.Although, in the case of the existing consortium on the property market, does not generate a strong grip to regulate consumer preferences, it can hurt small developers.Familial relationship is investigated through the rule of reason approach, based on the perceived impact that it is capable of damaging the business competition market.In court decision, the problem will be returned to the illegal per se approach within the law.This makes the corporate network of kinship tie very difficult to intervene.Another similar parameter of relationship is the bond of friendship, which is even more difficult to trace.The existence of third parties or intermediaries also makes it more difficult to identify the network.
Furthermore, regulations regarding spatial planning are not integrated with other laws such as Law 5/1999 regarding unhealthy business practices.Regulations on spatial planning are affected by the dichotomy from this business competition law, such as land possession and consumer housing.The impact caused by the developer network is still felt in different regulations.Law 26/2007 does not contain the role of inspectors in enforcing spatial planning regulations, such as controlling zoning regulations.On the other hand, the capacity of local government in managing spatial planning is still poor.The existence of a network of development companies that control extensive land reserves cannot be properly regulated institutionally.The regulations related to land and agrarian affairs do not enact rigid land bank regulations.This entitles big developers to freely expand the area as an investment asset.

Developers network analysis
The identification of prominent developers in large-scale landed housing both in terms of degree centrality and proximity centrality shows PT Intiland Development, a subsidiary of the Intiland Group, as the most central actor in the network based on direct relationships.PT Intiland Development owns a large-scale residential project, namely Telaga Bestari, that covers an area of 578 hectares in Tangerang.This company has 8 strong relationships (degree) with other major developers in Indonesia including developers from the Ciputra Group, Lippo Group, and Sinarmas Group.The higher the degree value in a developer means that the developer will find it easier to access information, knowledge, and other resources on the network inhabited by the developer [27].These companies are interconnected by developing identical products.Products in real estate include commercial property leasing, industrial building management, hotel operations, and sports facility operations.The company's portfolio includes townships & housing complexes for the middle to upper class, single-storey & multi-storey building developments, tourism & industrial complexes in the Jakarta-Bogor-Depok-Tangerang-Bekasi (Jabodetabek) area, Surabaya, and a number of other cities.Meanwhile, holding the most prominent development actor in the network with indirect relationships is PT Jababeka, having 17 degrees of relationship with other developers.PT Jababeka is developing a 100-hectare Terace Country in Bogor.The Jababeka Group itself is an integrated industry-based independent city developer.PT Jababeka Tbk.offers a variety of industrial real estate and complementary infrastructure and estate management services, which the company believes to be attractive to businesses engaged in light, medium and heavy industries.PT Plaza Indonesia was found to be the most central actor acting as an intermediary.This company is held by Franky Oesman Widjaja, the son of the founder of Sinarmas Group, Eka Tjipta Widjaja.Companies with the status of central actors are considered to have the power of access and control over resources inside the actor's network.Investigations related to prominent developers through an analysis of the degree of centrality in largescale apartment developers show PT Ciputra Property as the central actor in the network with direct relationships and PT Multi Artha Pratama in the network with indirect relationships.Meanwhile, the proximity centrality analysis shows that PT Ciputra Property is still the most prominent actor in networks with direct relationships.Networks with indirect links show PT Sumber Daya Nusaphala as the most central actor.PT Karya Supra Perkasa and PT Citra Benua Persada were found to be the most central actors in their roles as intermediary actors.
Through network coverage attribute, it is learned that large-scale site housing developers have the most relationships at geodesic distance 4 with a frequency of 506, which means that there are 203 relationships between actors that must first pass through 3 intermediaries on the direct network, and the indirect network having a geodesic distance of 4 with a percentage of 9.9% or a total of 1,788.In largescale apartment developer networks, direct link produces the highest frequency at geodesic distance 4 as well.Whereas with indirect networks, the largest frequency value is at geodetic distance of 1. Affordability measurement shows that 59 landed housing developers have a direct relationship to the total of 86 developers, and in indirect network there are 92 developers whose relationships sums up to 135 actors.The apartment developer network directly creates one large subnet with 34 developers through indirect relationship, and even though it only connects 26 companies, it formed 8 separate subnets.The existence of this large number of sub-networks is due to the role of intermediary companies that only connect a few companies.
In relation to the principle of intermediation, the analysis of intermediary centrality in the network of large-scale landed housing developers still shows PT Intiland Development as the most central actor in the network with direct relationships, and PT Jababeka in networks with indirect relationships.Meanwhile, the structural hole concept shows that PT Lippo Cikarang has the lowest value of resistance on networks with direct links, while PT Jababeka occupies an indirect network.The central actor in this concept is considered to be more flexible in controlling the flow of information on the network.Identification of intermediate attributes in the large-scale apartment developer network in Jabodetabek shows that PT Ciputra Property, which is a subsidiary of the Ciputra Group, as the most central developer, with the lowest constraint in the concept of intermediate centrality and structural holes in the network with direct relationships.These actors hold control over the interaction/flow of information, influence, and higher dependence on the network [33].Meanwhile, the network with indirect relations shows that PT Duta Pertiwi, which is a subsidiary of the Sinarmas Group, has the position of the most central actor with the lowest obstacle.
The final attribute regarding cohesion shows that the opportunity of interaction between actors in a network with direct relationships are lower than in a network with intermediaries.The network of largescale landed housing developers has a lower density of 2.3% than that of the network with an intermediate relationship that with the density of 5.4%.Meanwhile, large-scale apartment development networks based on direct relationships also show a lower density of 5.3%, compared to networks with indirect relationships with 6.1%.Another network cohesiveness parameter is the clique formed in the large-scale landed housing development network with a direct relationship, producing 15 cliques, with PT Bumi Serpong Damai, which is a subsidiary of the Sinarmas Group, filling the most cliques.Meanwhile, the indirect relationship is consisted of 35 cliques with PT Kemang Pratama filling the most cliques.In the large-scale apartment developer network with a network based on direct relationships, there are 4 cliques with PT Buana Surya Makmur and PT Multi Artha Pratama filling the most cliques.Meanwhile, the indirect relationship is consisted of 17 cliques with PT Karta Supra Perkasa (an IOP Publishing doi:10.1088/1755-1315/1318/1/01201616 intermediary company) filling the most cliques.These companies that fill the most cliques have an important role to establish cohesiveness in the network.
Identification of the characteristics of emerging networks shows that there are differences as well as similarities between site-type housing developers and large-scale apartments.Significant differences include number of subnets, power of central actors, geodetic distance, density and presence of third parties.Similarities can be seen in network characteristics between large-scale developers, both site and apartment types.Developers that own a parent company have a tendency to network with other developers.Furthermore, in networks with direct connections, the highest frequency of connection is at geodetic distance 4.This means that they can directly reach each other after passing 3 intermediaries.Whereas within the concept of density, the opportunity of interaction between actors in a network with a direct relationship is lower than in a network with intermediaries.Another characteristic is the clique created in networks with more indirect relationships than in networks with direct developers, indicating the role of intermediary companies in building network cohesiveness so that it is stronger.
Residential developers of both the site type and large-scale apartments create a social network.The existence of this network enables development companies to relay sensitive information related to the property market such as input-output and pricing.The formation of this social network raises the presumption that the property market in Jabodetabek is not running competitively.Although the evidence of this practice cannot be seen empirically, the indications that emerge can provide a picture of an anti-competitive market.These indications include high land and housing prices and land possession by a network of developers, which show a semi-oligopolistic structure.
With large housing backlog in Jabodetabek and increasing number of job seekers, high property prices are an obstacle for the public.The inability of the public, especially the middle to lower class, to acquire housing in Jabodetabek can create urban problems such as formation of slum areas, criminality, and health issues.Additionally, the network that was formed among the big developers strengthens company capitalization.This can make it difficult for small developers to compete in the property market, even if these small developers could provide better product quality.
The existence of third companies or intermediaries makes the network between developers difficult to detect, so that unhealthy business practices are difficult to observe.Intermediary companies strengthen the network between developers.This makes it difficult for the government and planners to act as regulators to determine policies.Land possession controlled by the developer network becomes difficult to manage, resulting in suboptimal development.High price and size of land banks hinders the developers from providing benefit to the public.

Conclusion
There are 1,161 landed housing projects spread across Jabodetabek, with the highest number of developed large-scale projects situated in the Bekasi area, accounting for 123 projects.In this type of apartment, there are 45 large-scale projects concentrated in South Jakarta.The largest housing developer group, the Sinarmas Group, controls 15.04% of the area of site housing projects in Jabodetabek, while the highest unit price that can be found among large-scale apartments is Anandamaya Residence, developed by PT Brahmayasa Bahtera for 60 billion rupiahs.
The second conclusion is that the land possession condition in landed housing in Jabodetabek is headed towards a relatively competitive market.The eight largest development groups control 53.95% of the total area of housing project in Jabodetabek, but the largest group of developer, the Sinarmas Group, still controls less than 20% of the land supply for housing.Based on the assessment of control of land supply for landed housing projects, a monopolistic competition-type market condition is obtained, namely a condition where there are increasingly more and more housing development companies in the property market so that consumers are provided with many alternatives.Meanwhile, the land development from the formed network shows that the land possession structure for site housing in Jabodetabek is semi-oligopolistic.
The third conclusion is that the central actor in the network of large-scale landed housing developers with direct relationships is PT.Intiland Development, which is a subsidiary of the large Intiland group, while the central actor in the network with indirect relationships is PT.Jababeka.In large-scale apartment development network, it was found that PT.Ciputra Property was the central actor in the network with direct and indirect relationships, while those with indirect relationships showed PT.Multi Artha Pratama and PT.Sumber Daya Nushapala as central actors.Residential developers who own a holding company have a tendency to network with other developers.
Actors in the residential development network are more difficult to interact with than actors in the apartment development network.The role of intermediary companies can build cohesiveness in the network, building it stronger.According to the all finding above, KPPU should be given stronger authority as a pioneer in enforcing the law against unhealthy business practices, as well as integrating its role with other legal institutions.The highlighting point is how to integrate policies related to the housing market and any business related to land and agrarian with spatial planning institutions.As government, conducting interventions that can be carried out as innovations in the stipulation of spatial control tools such as Zoning Regulation (PZ) is highly expected to control land banking activity which may threat the public right.In addition, PZ can be integrated in regulating the designation of zones or areas in accordance with the prohibition on monopolistic practices and unfair business competition as stipulated in Law no.5/1999, of course, accompanied by sanctions and firm action for violations.Moreover, designing a taxation system that is expected to be an incentive or a disincentive in the context of housing and land development.

Figure 1 .Figure 2 .
Figure 1.Distribution Map of Large Scale Landed Housing Projects in Jabodetabek

Figure 3 .
Figure 3. Sociograms of Social Network for Large Scale Housing Developers in Jabodetabek with Direct Relationship

Figure 4 .Figure 5 .Figure 6 .
Figure 4. Sociograms of Social Network for Large Scale Housing Developers in Jabodetabek with Indirect Relationship ) from the United States, classifying networks into three types.The assessment carried out in this section will cover inter-developer network with direct and indirect (intermediary) relationships.The assessment diagram used in classifying the network is shown in the following figure.

Figure 8 .
Figure 8. Network Diagram of Assessment Characteristics Based on Land Ownership Area of the Developers

Table 1 .
Large Scale Landed Housing Projects Area Classification in Jabodetabek 07% of land development area for large-scale landed housing projects is established by 31 projects with ≥ 500 Ha area and the land possession reaching 52,464 Ha.A significant number of large-scale projects is still occupied by of ≥100 to <200 type, reaching the less of projects, but only contributes in 7,801 hectares, or around 11.16%.

Table 2 .
Distribution of Large Scale Landed Housing Projects in Jabodetabek

Table 3 .
Land Acquisition over Housing Developers in Jabodetabek Land area was calculated from the area of existing housing projects that was being marketed by the developer companies.Collection of land area data was carried out by reviewing the compilation of company data, web scraping, magazines, and property news (2015-2020).Housing projects that are not listed in these data were explored by digitizing My Maps manually (2020) ].

Table 4 .
Relationships between Large-Scale Housing Developers