THE POLYVOCALITY IN ACTION: EMBODYING THE EPIPHANY ON PUBLIC HOUSING CONUNDRUM

Purpose of the study: This paper aims to elaborate on how the institutional logic of public housing provision transforms from the old logic into the new logic. By encapsulating tacit-knowledge from the shift, this study focuses on how to explicate a model of multi-criteria decision-making for executing official residence projects in Indonesia. This research also aims to recuperate the future of the public residency field. Methodology: The methodology employed was a mixed method. The qualitative method was firstly applied by utilizing semi-structured interviews to build a decision-making model. Later, a quantitative method was implemented to improve the consistency of the model by using the Analytical Hierarchy Process protocols. Thirteen project representatives and ten decision-makers were being involved as the main informants. Google Sheets web-based software was applied for analyzing survey results by making use of a mathematical model for the Analytical Hierarchy Process. Main Findings: The result indicates that land not used as paddy field, fair house price, safe from flood location, serene ambiance, and developer experience has played a significant part in affecting buying decisions for residency projects. To carry out the new method, the total scores of the AHP calculation should be above 50. Otherwise, the government shall opt for the old mechanism. Applications of this study: The paper contributes to the public sector accounting field in the area of asset management and public housing procurement. The paper also gives a strong basis for the field to make use of the model of Indonesia's public housing provision. The government can adopt the research framework for choosing the new model or the old one. Novelty/Originality of this study: This study might be the only one of its kind. The research was using a multi-method to achieve the objectives. To generate a multi-criteria decision-making model, grounded data of excessive interviews were abstracted. Then the abstracted tacit-knowledge was tested with AHP to provide a consistent model.


INTRODUCTION
Under President Jokowi's administration, the government of Indonesia has intensively conducted massive development in infrastructure (Negara, 2016;Warburton, 2016). Government infrastructure projects are designed to meet the socioeconomic needs of developing countries (Ahsan & Gunawan, 2010). Figure 1 shows that half of the government's total projects are construction projects. In this case, those projects comprise road expansion, building arrangement, and house provision 1 . To realize the projects, the government has to carry out the public procurement mechanism based on the related regulation 2 (Ray & Ing, 2016;Zukhrina-Oktaviani, 2015). Being involved in the praxis of the public sector area makes us 3 witnessed that over the past four years, the realization of the construction projects has been exhibiting a positive signal. It indicates that the country's economic growth is on track (Kaming, Olomolaiye, Holt, & Harris, 1997). However, of all government construction projects, especially those handled by the Ministry of Public Works and Housing (PUPR), the government encounters recurring predicaments, namely cost and time overruns, especially relating to the completion of housing projects (Table 1). Fortunately, the government is not alone in this regard because other developing countries are dealing with the same problems too, such as Southeast Asians: Malaysia (Chai & Yusof, 2015), Thailand (Toor & Ogunlana, 2008), Cambodia (Durdyev, Omarov, & Ismail, 2017), and Singapore (Hwang, Table 1 exposes a critical issue. The low completion rates on housing projects happen to be a burden in achieving the most challenging target of the "Nawa Cita": providing houses to the society, particularly officials and low-income citizens. However, the essential discourse in the public sector field is to discover a solution for handling the recurring problem of time overrun in housing projects (Arnaboldi, Lapsley, & Steccolini, 2015). Then, the next question is that "is there one appropriate solution to minimize the time overrun that has been a major obstacle in housing projects for 1 3 Using "Me" as the first person pronoun delineates the ontological belief employed (Kamayanti, 2016 The decision-making process entangles the appraisal of a multitude of both qualitative and quantitative judgments, including the assessment of a compound set of the interplay between these elements (Ball & Srinivasan, 1994). There is a scant amount of studies providing a framework for make-or-buy decisions (Sillanpää, 2015), even more in the public housing area (Palm, 2016). Serrano, Ramirez, and Gascó (2018) found that managers often face difficulties in the case of make-or-buy decisions. It becomes a strategic proposition, yet rhetorical. Hwang, Ko, and Goan (2007) faced a typical issue with the make-or-buy decision model in the field of cellular manufacturing and generated a multi-criteria model to help to fill in the void. McIvor, Humphreys, and McAleer (1997) attempted to unravel a panacea model for solving problems emanating from the make-or-buy decision in the private sector. Culled from a case study in a global telecommunications equipment company and its suppliers for the last ten years from the year 2000, Humphreys, Lo, McIvor (2000) uncovered make-or-buy decision problems affecting the profit margin of the company. In the meantime, Tayles and Drury (2001) buttressed with Sillanpää (2015) proposed a framework for the make-or-buy decision-making process. In sum, rather than creating the make-or-buy decision-making framework for improving the governance in public sector organizations, most of the previous studies played their roles by generating a framework to improve the supply chain efficiency and boost the profit margin in private sector companies.
On the contrary to previous studies, our study focuses on filling the research gap in the public housing field. Our research attempts to improve government judgment in the make-or-buy decision-making process, especially in the area of public housing since there is a scant amount of literature discussing the decision-making model for the public sector field. Therefore, the purpose of this study encapsulated succinctly from described phenomena, problems aroused, and praxis-research gaps is to answer three main research questions. They are (1) how far are the actors' roles in the Ministry of Finance in shifting the organizational logic in the public housing field in Indonesia? (2) how should the model of multi-criteria decision-making be befitted for public housing provision? (3) how should the future of public residency procurement be looked like?
To answer the first research question, we carried out the qualitative research method. However, the result of qualitative method employment in developing decision-making criteria always raises a common issue (Ball & Srinivasan, 1994). The assessment of qualitative criteria is naturally judgmental so that the major issue emerges and deserves attention is the aptitude to extricate consistent decisions (Ball & Srinivasan, 1994). To address the issue of inconsistency related to the qualitative criteria of the make-or-buy decision-making model, we employed the Analytic Hierarchy Process (AHP) that is widely applied for developing the consistency of a multi-criteria decision-making model (Saaty, 2004). Implementing the quantitative method of AHP is important to answer the remaining research questions. Then, we conducted the study at the mezzo level: the Ministry of Finance of Indonesia (MOF). The sole reason is that the MOF was appointed as the pilot project for developing and implementing the model of the make-or-buy decision in the sector of official residency in Indonesia 9 (Qadri, 2019;Qadri et al., 2020). Nevertheless, the results of this study are still relevant to other government institutions for being adopted in their housing projects because every government institution has the same regulatory framework 10 (Qadri, 2019; Udoyono, 2012).
Subsequently, the results of our study can be used as a basis for further research at the macro level using the nation-wide survey to develop a generic model in public housing policy for the government of Indonesia as well as other developing country governments that have a similar setting.

Meta-theoretical Framework of Public Housing: Providing Homogeneous Residency for Government Officials
Ruonavaara provoked an interesting discourse: "there is no theory of housing, but a meta-theoretical framework for various kinds of problem-solving analyses of housing procurement" (2018). His statement becomes a prominent proposition in the housing field of study, especially in the area of the public sector field (Kimhur, 2020). He adds another insightful discourse: "residency is a rather multi-faced theme" (Ruonavaara, 2018). Interestingly, Wang (2012) expressed approval to his proposition: "to provide public housing is to minimize differences in the built environment regarding location, physical quality, and size of the residence." He further explained that the most prominent characteristic of public housing projects is "the housing landscape was strictly regulated by a series of planning and building design guidelines" (2012).
Meanwhile, the objectives of the official residence program are to meet the necessities of basic shelter, to create the stability of living cost in the local area, to maintain the social stability, and to increase the integration of society as a whole (Field, 1987;Teo, 2015). In the context of Indonesia, the government defines public housing as: "a building that is owned by the state and functions as a residence for officials and a means to foster their family and support the settlement of their duties." 11 From this definition, two basic understandings are conforming with what Rohe (1995) specifies as the cooperative housing: first, the house is owned and managed by the government; second, the residents, who are officials and their family, are free to utilize the house by their duties and functions (Miceli, Sazama, & Sirmans, 1994;Rohe, 1995).
In this case, the primary hurdle confronted by the government is that the former residents deny to handover their usufructuary rights to the new residents (Lang & Novy, 2014). The formers are usually reluctant to be supplanted because they have lived in for age (Bockman, 2018). In the United States, there is a "prescription" often used to treat the reluctance of the old occupants who are misdirected in claiming a statehouse that is not theirs, namely: undertake a major renovation to the house, so that inevitably the old residents will shift to another dwelling. Thus their most notable claim to power will be deprived, specifically: their body inside the house (Bockman, 2018). Furthermore, Walker (2001) provoked an indigenous formula to overcome the reluctance problem of former residents, that is transfer the house management rights to the private party, limit all resources supplying the needs of the old inhabitants, isolate them from the surrounding environment, then as time goes by, the house will destroy its dwellers.

Previous Researches on Housing: Unveiling the Criteria of Make-or-Buy Decision in Residency Projects
A number of studies suggest that the most influential drivers in make-or-buy property decisions are physical quality problems and budget-related issues (Moschuris, 2007). Sillanpää (2015) argues that project managers decide to execute a buying decision because of the budget threshold problem, lack of knowledge, and lack of resources. We discovered that supplier credibility, budget sufficiency, and physical quality were come off second-best to the other salient factors in influencing the make-or-buy decision, especially those associating with the decision to buy a property. study that high quality of the living environment, in the form of high quietness level, provides more attraction to dwelling buyers in Geneva, Switzerland. Furthermore, a strong sense of neighborhood among dwellers in a community can foster a strong bond in the living environment (Li, 2008). Thus, the sense of neighborhood is one of the most important factors in defining the living environment (Grum, 2017).

METHODOLOGY
We employed a mixed-method in conducting the research. We used the combination of the qualitative and quantitative approaches to develop the multi-criteria decision-making model of housing provision. We embraced the abstractionistnarrative paradigm developed by Boje (Boje, 2019; Rosile, Boje, Carlon, Downs, & Saylors, 2013). Boje (2019) raises three philosophical questions regarding the use of his paradigm: "what is the nature of being-in-the-world?" (ontology); "does the researcher intend to unravel existing reality through the story or develop a change in dominant reality?" (epistemology); "how does the researcher intend to use data that are gathered from stories?" (methodology). Rosile firmly argues that the researcher should iteratively ask himself to answer the questions in order to: "create robustly, and rigorous storytelling research that matches ontological and epistemological assumptions with appropriate methods" (2013, p. 12). Responding to Rosile's argument, we viewed the nature of being-in-the-world as constituting reality. Therewithal one of the goals of our study was to construct a change in the reality of the public procurement field, specifically in the area of public housing provision.
The paradigm of abstractionist-narrative is built on the basic assumption that the researcher is in the necessity of abstract concepts for utilization in the upcoming generalized study (Rosile, Boje, Carlon, Downs, & Saylors, 2013). Overlying on this assumption, the paradigm epistemology that should be implemented is (1) to discover for linguistic schemes, and then (2) to abstract them into the transcendent level of knowledge (Rosile et al., 2013). This epistemology notifies a methodological purpose in extracting what has been uncovered (Heracleous & Barrett, 2001). The reason of why we used a set of Boje paradigm and epistemology was that the nature of our research was aligning with what Rosile suggested that if a study is intended to alter sets of action, at that point exploiting Boje set of paradigm and epistemology is highly preferable (2013, p. 10).
The method to uncover linguistic schemes was gained by implementing the Boje qualitative method of storytelling organization (Boje, 1995(Boje, , 2019. Aligning the method with the set of paradigm and epistemology proposed by Boje was mandatory for setting the tone of the paper and adjusted it with the paradigm-epistemology (Boje, Rosile, Durant, & Luhman, 2004). Meanwhile, the method to abstract the schemes into the transcendent level of knowledge was achieved by studying multiple discourses that can uncover significant elements affecting change using the analytic hierarchy process (Saaty, 1982(Saaty, , 2004. The quantitative method of AHP was applied to build a consistent set of criteria that can be utilized over a phenomenon with theoretically typical contextual measurements (Ball & Srinivasan, 1994;Rosile et al., 2013;Saaty, 2004).
We utilized the combination of a qualitative and quantitative method to develop the multi-criteria decision-making model in the public housing field. This research followed the mixed method procedures suggested by Boje (2019), Saaty (1990), and Saunders (2015) for executing the research steps of data collection and findings analysis. The data collection and analysis steps were embedded in three important stages of the research method, namely (1) stage of gathering terse stories in the Ministry of Finance of Indonesia, (2) stage of conducting a systematic review on the house provision field, and (3) stage of implementing the AHP. We made use of the data collection and analysis for the qualitative method of the Boje storytelling organization framework in the first two-stage (Boje, 1991(Boje, , 2019. After that, in the last stage, we implemented the data collection and analysis for the quantitative method of the AHP (Saaty, 1982(Saaty, , 1990.

Stage 1: Gathering Terse Stories in MOF
We started our research by carrying out the participatory observation at the beginning of January 2019 at the head office of the Ministry of Finance until July 2019. We devoted our time to attend several meetings that discussed the residency provision. We took some field notes and actively participated in some conversations between colleagues in the office. We gathered stories from project managers, procurement committee, and vendors by recording and transcribing every important meeting. In general, we have outlined four of great importance events for depicting the stories of public housing perplexity. Those events were (1) meeting with National Asset Management Office (DJKN) on January 18th, 2019; (2) meeting with MOF's Internal Auditor Office on January 25th, 2019; (3) workshop on Public Housing Procurement with PUPR on January 29th, 2019; and (4) meeting with LKPP on February 13th, 2019. Eventually, we condensed the storied discourses and put them chronologically in a memo. Afterward, we examined them with thematic analysis techniques. We opined that stories, in the MOF head office, were not simply something which individuals advise to others to engage, nor simply something they do when they impart. Rather, stories are the veins through which changes throb in the core of MOF life.

Stage 2: Conducting Systematic Review on House Provision Field
Along with our observation, we performed a systematic review to assemble the best practice puzzles in the field of residence provision (Grum & Kobal Grum, 2015). We searched for articles concerning the field to refine the storied narrative from our observation. We opted for Emerald, JSTOR, Science Direct, and Google Scholars as the publication places for identifying relevant works. We set 20 years: 1999 to 2019 as the period of the systematic review. Then we picked several keywords, namely house selection, site location, store location, and analytic hierarchy process, as the foundation of our searching. We entered those keywords interchangeably into the publication places. Finally, we discovered 346 articles discussing the housing provision, but only 30 articles were relevant to the storied narrative. The results of the systematic review substantiated the stories we collected during our observation. We developed the preliminary criteria and sub-criteria for the inputs of the following stage.

Stage 3: Implementing The AHP
We employed the AHP to sharpen the judgment and generate consistent decisions (Ball & Srinivasan, 1994; Schniederjans, Hoffman, & Sirmans, 1995) for the government, especially MOF on residence provision cases. The AHP was developed by Saaty (1977Saaty ( , 1982 on the principles of trade-off and empowered the decision-makers to improve the trade-off verifiably through organizing and scrutinizing a set of mutual pairwise comparison matrices (Ball & Srinivasan, 1994). The AHP can help decision-makers in attaining a consensus over critical criteria. Even though there is always an opposition against the consensus, decision-makers can still pay more attention to these areas of disagreement to reach the consensus (Ball & Srinivasan, 1994).
In conducting the AHP procedures, we tested the primary criteria and sub-criteria resulting from storied narrative and systematic review by conducting the preliminary survey to 13 project representatives. Of those participants, ten persons only have given relevant answers. Afterward, we continued the AHP procedures by conducting an extended survey based on a 9-point scale measurement to ten strategic project decision-makers, but we excluded three of them because they lack the experience to understand the survey questions. The main reason why we chose a small sample in the form of ten strategic project decision-makers was based on their expertise consideration. The same reason is expressed by Lai, Wong, and Cheung (2002) as well as Russo and Camanho (2015). Their arguments are supported by the founder of the AHP, Saaty (1990Saaty ( , 2004, who argued that the AHP could be executed with small numbers of respondents who have the expertise to cause minor errors in judgment. The project decision-makers involved in this research had an average of five years of construction project experience. The participants were provided with a background of the first and second surveys, problems in multi-criteria decision making, and a brief explanation about the AHP method. We required the participants to answer the AHP based questionnaire carefully as they had to prioritize their answers first before they filled in the questionnaire. We conducted a meeting with a majority of participants to build a consensus about decision-making problems using the AHP. We gave a chance for the participants to evaluate and revise their answers. The final answer became the input for the AHP procedures of analysis to compose the final model of the multi-criteria decision-making model for public housing provision in Indonesia. To achieve the consistency matrix of the final model, we tersely summed up the AHP procedures into six primary steps (Table 2). importance between two choices of activities. The measurements are as follows: One means equal importance or two activities contribute equally to the objective. 3 means weak importance of one over another or experience and judgment slightly favor one activity over another. 5 means moderate importance or experience, and judgment strongly favors one activity over another. 7 means strong importance or activity is strongly favored, and its dominance is demonstrated in practice. 9 means absolute importance, or the evidence favoring one activity over another is of the highest possible order of affirmation. 2, 4, 6,8 mean intermediate values between two adjacent judgments when compromise is needed. The measurement results are reciprocal, meaning that if activity i has one of the above nonzero numbers assigned to it when compared with activity j, then j has the reciprocal value when compared with i.

STEP 3 Composing Pairwise Comparison Matrix (PCM
The PCM shows the extent to which one criterion is dominant against another criterion of the same level. The PCM is presented as follows:

STEP 4 Measuring Priority Vector Normalization
We measured the vector normalization by summing each number in the same column of PCM, then dividing each number with the total sum, after that summing the result of the division in each row to determine the geometric mean of the row.

STEP 5 PCM Consistency Test
We carried out the consistency test using a consistency index, random index, and consistency ratio for each PCM. There are 7 PCMs for the main criteria and 24 PCMs for sub-criteria. The consistency index (CI) and the consistency ratio (CR) is computed with the following formula: RI means random index and is measured as follows:

RESULTS AND DISCUSSION
To implement Boje's principles (Boje, 1991), we picture the MOF as the storytelling organization because the MOF has terse stories to tell, the stories of its people undertook a process of enriching the organizational logic of public house provision by developing an indispensable framework of make-or-buy decision. The people took part in a broad type of stakeholder groups, i.e., project managers group in DJKN and MOF Assets Management Centre (Romadan), to march in the procession of information and manage collective memory of the organization, the MOF, utilizing storytelling (Boje, 1991). Old stories of changing organizational logic in housing provision are reminisced in the collective dialogue among stakeholders as they compelled the implementation of the buy-directly-from-developer mechanism, then unfolding stories about composing the framework of make-or-buy decision is created, and future stories on public housing management are shared among stakeholders to make sense of recent dynamics in the organization.

The First Objective, Precedent Stories: Shifting The Logic of House Provision
In this section, we describe how the face of the housing provision mechanism in Indonesia is trying to be altered into a new shape. We present the findings and analyses of our study following Boje (1991) style in presenting the results of his study. There are several plots depicted to answer the first objective of this study. Those plots are terse story 1 about the initial trigger of mechanism alteration, terse story 2 about the reason for mechanism change, terse story 3 about raising internal supporters, and terse story 4 about pushing the start button for mechanism shifting.

1.
Bro, do me a favor, will you? DJKN has an issue with

2.
their construction projects in their local unit.

3.
They ask us to join the meeting at 10.00 a.m. in their office.

9.
Here is the DJKN's notice, learn about it first 10. before you go there.

11.
Ask Munir or Tedjo with you because I cannot accompany you. In Terse Story 1, Hans, the supervisor of the procurement committee, is enjoining his staff to attend the meeting discussing the procurement problem in DJKN (line 2-4). From another source, 12 We have found that the problem is about the housing procurement for officials in Mamuju, West Sulawesi. Mamuju's project manager is questioning the criteria and procedures he should perform if he wants to execute the budget of official house provision via a buydirectly-from-developer mechanism. The meeting Hans 13 asks Ari to attend is going to discuss further the mechanism. Henry 1.
About our project in Mamuju, we do not know whether

2.
the project manager could apply the buy-directly-

4.
Although it is legal based on the regulation,

5.
I have not found any institution carrying out the mechanism.
6. This is why your role here becomes important so that  Terse Story 2 tells you more about the meeting at DJKN head office. Ari, Munir, and Ron represent Romadan in the meeting, while Henry, chief coordinator for strategic projects, represents DJKN. Henry relies heavily on Romadan's resolution regarding the housing provision policy (lines 6-8). He figures out that there is no institution operating the procedure of buying mechanism, whereas, by regulation, it is allowed (line 4-5). The mechanism was not the institutional logic that prevailed in the field of residence provision, but the build method was. Henry persuades Ari and his team to give DJKN support for breaking the logic by executing the direct buying mechanism (lines 9-10). He lets Johan in to provide field experience insight in respect of house provision project in Papua (line [14][15][16]. Johan unveils one strong reason to implement the buy-from-developer method in Terse Story 2 (line [19][20]. Ari explains that the Mamuju case is a typical "go-or-no-go" case that cannot be solved without involving the other stakeholders in the decision-making process (lines 22-24).

1.
There will be 2 projects in Mamuju for piloting the direct 2. buying mechanism.

4.
Another one is for DJKN.

5.
Implementing the buy-directly-from-developer method is not

8.
Sadly, the regulation does not provide the criteria on how we 9. should execute a build or buy mechanism.

10.
I personally face this problem for the first time.

12.
this case is related to LKPP authority.

13.
As long as I concern, PUPR has no authority in regulating 14. the procurement procedure.

15.
PUPR regulation is only pertinent to technical 16. details of the house.

17.
If we want to do the direct buying things, the project

19.
It is important because the stock will guarantee the project 20. the manager that he will get the house he needs.

21.
We should put it on the requirements list of developer criteria. After the meeting, Hiro ravels a background story of the selection of two pilot projects to Romadan representatives. 15 He presumes that MOF can be the first ministry to implement the scheme, but MOF needs to develop the criteria for determining which project should be executed via build mechanism or buy-directly-from-developer method.
There are so many local offices requesting the implementation of a direct buying method. 16 MOF cannot grant any of their requests before resolving the criteria of make-or-buy decisions. 17 To bolster the criteria development, MOF has selected two projects under DJPB and DJKN authority as the pilots. 18 The most important aspect that has to be considered in developing the criteria is the availability of house inventory. Hiro insists that the housing developer must have sufficient house inventory to complete the project (line 17-18) so that it is expected to mitigate the risk of delay. In the meantime, Andre suggests Romadan communicate with LKPP regarding the procedure of direct buying scheme because it is still within LKPP functions (lines [11][12]. Despite Romadan's attempts to shift the organizational logic by implementing a direct buying scheme, MOF and other institutions still have to follow the house technical specification rule as mandated in the PUPR regulations 19 (line 15-16).

1.
Essentially, our regulation has provided the institution

4.
I am not mentioning one type of procurement, not goods,

5.
services, nor construction projects, but all of them.

6.
I am saying that government procurement will eventually 7. follow the best practices.

8.
We do not want to create rigid policies nor procedures.

9.
So, in the case of MOF official house procurement, I hope that 10.

11.
buying mechanism and, further, the role model.

12.
In order to develop the direct buying criteria, you can adopt 14 As stated on the MOF annual budget year 2019, only Mamuju Treasury Office and Mamuju Asset Management Office will implement the direct buying mechanism. The information has been confirmed by Hiro in post-meeting conversation on January 29, 2019. 15 The Itjen and Romadan team had an extended conversation after the meeting on 29 th January 2019. We didn't record the conversation but We memorized it, then We wrote the critical points in our field notes. 16 Ibid. 17 Ibid. 18 Ibid. 19 Loc.cit. Number 4.  Terse Story 4 describes the final story of the institutional logic change in the public housing field. Not long after the meeting with Itjen, Hans and Ari decide to visit the LKPP office at Kuningan, Central Jakarta (Terse Story 4). Their objectives are to discuss the MOF plan on establishing the internal regulation on official house provision and to gain information about LKPP experience in solving the housing problems. The chief of construction procurement policy, Ardy, and the chief of construction services policy, Sean, welcome Romadan representatives. Since Hans and Ari have expressed their interest in attempting direct buying implementation in MOF, Sean is delighted to support MOF and challenges them to make MOF a role model (line [9][10][11]. He unveils that the procurement regulation is designed to adapt to the best practices (lines [1][2][3][4][5], in terms of the regulation does not arrange the details of procurement procedures. 20 It only states the principles that have to be allowed in the 3 general procurement processes (lines 6-7; 12-15). Those processes are procurement planning, vendor selection, and contract execution (line [16][17][18][19]. Based on Terse Story 4, Sean of LKPP appends that the MOF representatives, Hans and Ari, have to look at the best practices in the field of residence provision (line [12][13]. Afterward, he suggests them to adopt the principles of best practices for creating the basic criteria as well as procedures (line [12][13]. Sean also tells the Romadan team the way they should develop the basic criteria of make-or-buy decisions for house procurement (line [14][15]. He firmly states that if the decision is to buy residences directly from the developer, then MOF has to develop the buying procedures because they have not been ratified in any related regulations (lines [16][17][18][19][20][21][22][23]. Furthermore, MOF has to enact an internal decree that has substance on the essential criteria of make-or-buy decision and direct buying procedures (lines 24-28). In sum, LKPP has the same understanding of MOF about the direct buying mechanism. LKPP also supports MOF in stipulating the internal decree on the multi-criteria decision-making and the direct buying procedures of house provision.

The Reflection of Precedent Stories: The Key Roles of Key Actors
What can be inferred from the precedent stories is that the dynamics of shifting organizational logic inside government institutions involve both internal and external stakeholders. In the case of MOF, there are several actors embroiled. The main actors are Hans and Ari of Romadan. Their roles are gathering every support they need from internal and external stakeholders to develop and implement the multi-criteria decision-making in the field of house provision. After the criteria are established in MOF regulation, direct buying is eligible to be implemented widely in MOF. However, the main actor roles are affected by supporting actor behavior. The supporting actors are Henry and Johan of DJKN. They have played their roles in influencing the Romadan team to start the initiation of a direct buying scheme in MOF, beginning with the Mamuju project.

The Second Objective, The Unfolding Stories: Constructing Criteria of House Provision
After the LKPP meeting, Hans of Romadan asks us to help him develop the multi-criteria of make-or-buy decision. To fulfill his request, We need some insights from international best practices in the field of residency provision. Then we conduct a systematic review and use the review result to refine the storied narrative from precedent stories. The combined results of storied narrative and systematic review results are confirmed to Bob, a procurement expert of Romadan, who has more than eight years of field experience in the construction project. Bob accentuates two criteria recommendations to strengthen the make-or-buy decision model. Firstly, the unit must have no idle land asset based on the unit's statement of financial position and notes to financial statements. Secondly, the total cost of the housing project is sufficiently funded by the government, as stated in the unit's budget execution notice.
Two things are essential before you perform the buy-directly-from-developer mechanism! First, you have to make sure whether the unit that proposes the execution of the mechanism has an unused land asset in its working area. If the unit does not have any unused land, then you, as the asset manager of the ministry is obliged to locate an unused land in the other MOF units in the same working area. If there is no idle land asset in the area, then you can go on executing the mechanism. Second, the total fund of the house provision project has to be established in the unit budget. -Excerpt from an interview on August 23rd, 2019.
Then combined results of storied narrative and systematic review are incorporated with the meanings from the expert interview to generate the preliminary criteria decision-making of house provision (Table 3). We ran the six-steps AHP procedures to produce a consistent multi-criteria so that the MOF decision-makers can use them as a tool for generating consistent decisions. We presented the results of the procedures in Table 4-6 and Figure 6-9. What can be concluded from the AHP results are: first, to run the direct buying mechanism, the total scores of AHP calculation must be above 50. Otherwise, MOF strategic decision-makers shall select to implement the build procurement method. Second, We found exciting facts that amongst 7 criteria, the project decision-makers emphasize living environment, budget sufficiency, and land availability as the top 3 criteria with the highest scores. If the total scores among the top 3 combined, the results are 60,78%. So, this means that there is a way to increase the possibility of a project to be executed with direct buying, that is, by maximizing the scores within the top 3 criteria. Meanwhile, we discovered the big 5 of sub-criteria of make-or-buy decisions that are pivotal in enhancing the buy-directly-from-developer decision for the official residency project. Those big five are land not used as paddy field, house price stated at fair value, safe from flood location, lush and serene ambiance, as well as developer experience. Based on the preliminary criteria, We conducted the first survey to 13 project representatives using a google form to test whether the criteria were matched with their experiences ( Table 4). The criteria for selecting the project representatives are explained in the methodology section. The criteria proceeded into the next steps of AHP are the criteria that have 3points value of geometric mean and above (Saaty, 2004). The preliminary criteria are summarized in a 3-level hierarchy model ( Figure 6  The unit has no idle land asset 2,82 X12 Land area complies with the Standards 3,23 X13 The land is not used as paddy field nor a swamp 3,00 Humanities & Social Sciences Reviews eISSN: 2395-6518, Vol 8, No 4, 2020 House price is stated at fair market value 4,78 X22 House price is cheaper than total cost of self-constructed house 4,37 X23 House price is non-volatile 4,04 X24 House price covers purchase price, notarial fee, taxes, and excess of land 4,57

X3
Site Location 4,78 X31 Site location is relatively safe from flood and earthquake disaster 4,65 X32 Site location is nearby the work office 4,68 X33 Site location is closed to worship place 4,47

X4
House Physical 4,78 X41 House area complies with minimum house area based on related regulation 4,54

X42
House complex has a water network channel connected to water supply network system 4,54

X5 Living Environment 4,68 X51
The living environment is safe from social conflict 4,78 X52 House complex is placed in the one-gate system or cluster system 4,54 X53 Quietness level is relatively high 4,68 X54 House complex has paved road 4,68

X7 Developer Credibility 4,68 X71
The developer has ready-stock houses 4,54 X72 The developer has a healthy financial performance 5,00 X73 The developer has three completed project experiences 4,89 X74 The developer is a corporate legal entity 4,65 Source: Adapted from Decomposition Result -AHP To weigh which criterion is more important than another, we underwent the second survey to 10 strategic decisionmakers. The criteria and sub-criteria from the preliminary survey were being tested by applying AHP-style questionnaires with the 9-points scale measurements adopted from Ball and Srinivasan (1994). The survey results were used to compose the PCM for the main criteria of the decision-making model for residence provision (Figure 7). Using the PCM data, we exercised the priority vector normalization and consistency test procedures to calculate the consistency ratio of the main criteria ( Figure 8). For each main criteria, the consistency ratio must be above 0,10 to be accepted as the final main criteria (Saaty, 1982). We used the steps of calculating the consistency ratio of the main criteria to compose the consistency ratio for subcriteria. For each sub-criteria, the consistency ratio has to be above 0,10 to be approved as the final sub-criteria (Saaty, 1982). The conclusion of the PCM consistency test for sub-criteria 1 to 7 is that all seven sub-criteria are consistent because their consistency ratio is beyond 0,10 each. It means that the model can be implemented widely to provide the best decision (Ball & Srinivasan, 1994) in the field of public housing provision in Indonesia. If the consistency is poor or below 10%, supplementary information is needed to trace where the inconsistent information is derived from (Schniederjans, Hoffman, & Sirmans, 1995). The researchers should make sure that the expert as a research informant who provides that kind of information understands how to answer the AHP questionnaire (Saaty, 2004). The expert should build a prioritization of their preference before answering the questionnaire (Lai et al., 2002). The process of revising the answer should be conducted iteratively until the informant provides a consistent answer (Mani, Agrawal, & Sharma, 2014). From there, the consistent model of multi-criteria decision-making is constructed (Obeidat, Qasim, & Khanfar, 2018).
Rectifying the consistency ratio does not mean gaining the obvious answer to the issue being researched (Ball & Srinivasan, 1994). We agree, to some extent, with Ball and Srinivasan's (1994) argument saying that "the improving consistency ratio only indicates that the ratio is getting closer to being logically connected than to being randomly picked out." Based on the results of the consistency test for the main criteria and sub-criteria, the final weights of the multi-criteria decision-making model for residence provision are shown in Table 5 and summarized into the hierarchy in Figure 9. The land is not used as paddy field nor a swamp 8,33%

X2
Budget Sufficiency 20,52% X21 House price is stated at fair market value 9,84% X22 House price is lower than the total cost of the self-constructed house 1,84% X23 House price is non-volatile 4,55% X24 House price covers purchase price, notarial fee, taxes, and excess of land 4,29%

X3
Site Location 14,53% X31 Site location is relatively safe from flood and earthquake disaster 6,34% X32 Site location is nearby the work office 5,20% X33 Site location is closed to worship place 2,98%

X4
House Physical 5,06% X41 House area complies with minimum house area based on related regulation 1,76%

X42
House complex has a water network channel connected to water supply network system 1,02%

X5 Living Environment 21,92% X51
The living environment is safe from social conflict 3,50% X52 House complex is placed in the one-gate system or cluster system 1,30% X53 Quietness level is relatively high 12,97% X54 House complex has paved road 4,14%

X7
Developer Credibility 13,25% X71 The developer has ready-stock houses 1,73% X72 The developer has a healthy financial performance 3,94% X73 The developer has three completed project experiences 4,79% X74 The developer is a corporate legal entity 2,79%

Source: Adapted from AHP Results
A preview of the final factor weights displayed in Table 5 was synthesized from the PCM priority and resulting proportions of decision-making determinants like land availability, budget sufficiency, site location, house physical quality, living environment, community facility, and developer credibility. Generally, the results propose that the project managers in the public sector field consider the living environment as the first determinant for implementing the buying mechanism. Coming up into the second position is budget sufficiency. The third place is occupied by land availability and house location as the fourth rank. Developer credibility takes the fifth position. Community facility and house quality are placed in the last two positions.
Placing the house quality as the less important determinant is opposing the findings of Ismail  Our findings suggest that the living environment is the most attractive trigger in affecting the buying decision of public houses in Indonesia. There is a strong relationship between the living environment and house consumption by home buyers, especially the old ones (Wu, 2010 (2010) and also Ismail and Shari (2019) as emphasized by a "baby boomers" project manager at MOF of Indonesia, from the interview on 9 th November 2019, stating that "the living environment is the key factor to execute the buying mechanism on public housing project … and a cluster housing with one gate security is more suitable for public residences." His argument is strong evidence to support the relationship between the living environment and the implementation of buy strategy for public residency, as Bender et al. (2000) highlighted that a high level of environmental quality supplies more allure to home buyers.
We found that the most pivotal determinant constructing the criteria of the living environment is high quietness level, followed by a paved residential road, safe from social conflict, and one-gate system neighborhood. The level of quietness is the top-ranked determinant of the living environment because most project managers we have surveyed argued that a high level of quietness is important for the employees to fresh up their minds after spending most of their time at the office. This argument is aligned with Bender et al. (2000) study eliciting the fact that home buyers are leaving their apartment to buy a house that has a high level of quietness. "People choosing to buy a residence are normally going away from a flat where the loud sound sources are notable. If they opt to buy a residence, which signifies a large financial commitment, they desire to be in a quiet neighborhood" (Bender et al., 2000). The final model of multi-criteria decision-making for public dwelling presented in Figure 9 proposes seven determinants to execute the buying mechanism for providing public housing. They are (1) land availability, (2) budget sufficiency, (3) house location, (4) house physical quality, (5) living environment, (6) community facility, and (7) developer credibility. The main objective of the final model is to define logically 'which option to carry out' when project managers have a number of public dwelling projects to execute (Level 1 of Figure 9). In making the decision to run the buy a residence mechanism, project managers should perform multiple sets of procedures (Level 2 of Figure 9) by evaluating the makeor-buy criteria until arriving at the final decision (Level 3 of Figure 9). In order to execute the buying decision, the cumulative scores should be above 50% (Ball & Srinivasan, 1994). Otherwise, the public sector project managers have to carry out the make a dwelling procedure (Chauhan et al., 2008). However, project managers should pay attention to developer opportunistic behavior (Serrano, Ramírez, & Gascó, 2018). Using our final model for decision-making in public housing provision can reduce the risk of opportunism by providing a fair judgment on every vendor proposal. Thus, our research is consistent with the transaction cost economics theory, which evokes a proposition that the make-orbuy decision has a strong relationship with the degree of opportunism surfaced (Dibbern, Goles, Hirschheim, & Jayatilaka, 2004). Therefore, our research basically is strong evidence that the transaction cost economics theory does exist.

The Third Objective, The Upcoming Stories: Predicting the Future of Public Residency Provision
Based on the precedent stories, the government limits the mechanism for procuring state houses only in four ways, namely: build, exchange, grant, or buy-directly-from developer. 21 As explained earlier in this paper, the provision of official residence is carried out mostly through build mechanism. This study proposes the government with a framework to decide whether a project shall be executed with a build mechanism or direct buying method. Even so, the four methods have long been abandoned by most developed countries as they utilize the concept of public services partnership in managing housing projects (Morrison, 2016). The partnership framework is a breath of fresh air adopted by England (Mullins & Moore, 2018) and several commonwealth countries, such as Northern Ireland (Muir & Mullins, 2015), Scotland (Gibb, 2003), and Australia (Nygaard, Berry, & Gibb, 2008). The framework overlies in the form of a hybrid financing scheme: the provision of public housing is handed over to non-profit organization in the real estate sector (Blessing, 2012), then the total costs of provision are divided proportionally between the government and the nonprofit organization (Moore, 2018).
The non-profit organization then conducts the procurement mechanism by carrying out two main principles: considering value-for-money in price negotiations and implementing a design-and-build contract (Goodchild & Chamberlain, 1999 . However, MOF project managers have not captured the issue. They still contend that the implementation of a buy-directly-from-developer mechanism should immediately be performed as the alternative solution to improve the efficiency in the period of the official residency provision as well as boost the budget realization.
Maintaining the effectiveness of housing projects is also crucial in managers' opinions. Only by making sure that the project's outputs have met the minimum public residence standards can the public project managers implement the buying mechanism. In the future: "the official house management shall follow principles of the international best practices in the public services partnership."stated by informant Zack, Chief of MOF Procurement Centre. 22

CONCLUSION
This research aims to elucidate the dynamics of change in the organizational logic shifting of public dwelling provision praxis in Indonesia, and compose multi-criteria decision-making for public residence project using the versatile approach of the analytic hierarchy process as well as explicate the future of public residency outsourcing. The results unveil interesting findings that several actors play their part in shifting the institutional logic of public housing provision in Indonesia. The most influential actors are coming from the MOF because the MOF has been appointed by the National Procurement Authority as the pilot ministry for implementing the new logic of the buy-directly-from-developer mechanism. Then the MOF is urged by the authority to start the implementation by composing the multi-criteria decision-making model of the make-or-buy decision on public residence provision.
We helped the MOF through this research to compose the criteria. We found that the most important determinant among MOF project managers in executing the buying mechanism is the living environment, followed by budget sufficiency, land availability, house location, vendor credibility, community facility, and house quality. The government can run the buy-directly-from-customer method if the total scores of AHP calculation are above 50%. Otherwise, the government shall opt to exercise the build mechanism. Our study also reveals that there are five criteria for the buy-or-build decision that play major roles in influencing buying decisions for the official residency project. Those five are land not used as paddy field, house price stated at fair value, safe from flood location, lush and serene ambiance, as well as developer experience. Previous studies in the area of official residency provisions are infrequent. However, the future of public housing provision is bright. In the future, public sector project managers should think creatively to fund their projects because the government budget will be limited. To tackle the upcoming issue, implementing the public-private partnership scheme may be the best alternative solution.

LIMITATION AND STUDY FORWARD
This study might be the only one of its kind; notwithstanding, the paper provides several implications to both theory and praxis. Firstly, the paper contributes to the public sector accounting field, especially in the area of asset management and public housing procurement. The paper gives a strong framework for the field to make use of the model of Indonesia's public housing provision. Secondly, this study refines the transaction cost economics theory by providing evidence supporting the postulate. Thirdly, this research determines a framework for the government of multi-attribute criteria of the make-or-buy decision in official housing provision. The government can adopt our research framework for exploiting the option. The framework was abstracted from grounded data of excessive interviews. However, the efforts to contemplate the meanings of those data were relentless and time-consuming. So this research has not covered the issues related to the framework's implementation. Thus, further research should carry on the mission to scrutinize the implementation effect after the framework is being implemented widely. Last but not least, we evoke you with an idiom: 'buy thou house or continue thy dream then thou hast only thyself to blame.'