Post-Conflict Land Administration in Edendale N Wirewall Project

The Edendale N Wirewall case is instructive for land tenure administration, project management and group tenure schemes in post-conflict land tenure administration and housing project management. The Greater Edendale Area experienced major localised conflict in the lead up to the end of apartheid. The Edendale N Wirewall project is a problematic state subsidised housing project. Innovations in development techniques and land tenure models are necessary in times of far reaching change. However, innovations should be justified by robust empirical evidence if failure might have a significant impact on the intended beneficiaries. administration and project governance as new politicians come into power, the state is fragile and personnel changes and power shifts occur in the civil service. Fixing major problems in housing projects arising out of poor project management and poor project governance can result in costly remedial action and tenure insecurity in the decades following project completion. If the situation allows it, external expert checks on land administration operations management and far more frequent and independent project audits, on-site inspections and audit surveys may mitigate some of these challenges.


Introduction
Edendale N is a suburb in the Greater Edendale Area (GEA), which is some 15 km south west of Pietermaritzburg in KwaZulu-Natal. Both Edendale and Pietermaritzburg are part of the Msunduzi municipality. Edendale was a flash point in a localised violent conflict in Kwazulu-Natal in the 1980s and 1990s in the period leading up to full democracy in 1994 (Mkhise 2015). The article examines best practice in land tenure information management during far reaching change and postconflict land tenure administration project and operations management. It also examines the risks associated with innovations in group tenure forms and housing construction in the state-subsidised Edendale N Wirewall housing project ("The Wirewall case"). The Wirewall case is a regularisation and redevelopment programme to rectify a number of problems in a state-subsidised housing project in the Edendale N suburb where wire walls were constructed by spraying concrete onto a wire mesh.
The Wirewall case was studied during a project to regularise tenure, and to rectify legal boundary layout problems due to communication problems in land tenure information management and poor project management when the houses were constructed. The regularisation project was establishing who was in living the free standing homes and what should be done about those who had acquired their homes in what appears to be fraud scheme driven by a local politician. The regularisation project was also attempting to resolve problems that had arisen in a form of family tenure where, instead of individual families being allocated a single subsidised house, four families from the same extended family were allocated a larger house which they held in joint ownership.
The article should interest land information systems project managers, land professionals and researchers who work in post-conflict situations and similar volatile, uncertain land tenure administration situations. It contributes to the empirical work on best practice in project governance, risk management, communication procedures, document change procedures in land tenure projects and land administration systems, and on group tenure systems in urban areas.
The paper proceeds with review of the literature and general recommendations which are informed by the author's land administration and project management experience. Following this is a brief history of Edendale in the 1980s and 1990s, and then a description of the research method.
It then describes state subsidised housing and land governance in the Msunduzi municipality followed by a description of the Edendale N case and the analysis thereof.

Literature Review and Previous Work
Much of the literature on post-conflict land administration and project management is written from the perspective of international agents who assist in peace building and reconstruction. Of relevance to this discussion, the following can be expected to impact governance of development projects and the land administration system in post-conflict situations. Typically the state is fragile as the civil service changes and a new order is negotiated in a number of cycles. Internally displaced people invade vacant land which is often owned by people who themselves have had to flee.
Powerful factions grab land, corruption and intimidation are ubiquitous, crime rates are high, there is general maladministration and land administration systems are dysfunctional (Augustinus and Barry 2006, Barry 2006, UN-Habitat 2007, Todorovski et al 2016. In some conflicts reconstructing the records is an immediate and major challenge, as the buildings housing the records may be bombed or burned (Barry and Bruyas 2009) or removed deliberately by factions involved in the conflict (Todorovski et al 2013). Land administration systems and development projects may be compromised due to newly appointed, inexperienced or poorly qualified staff in state organisations, and due to the factors mentioned above (UN-Habitat 2007, Earnest 2015, Todorovski et al 2016.
In the author's observation, South Africa is a mild form of post-conflict situation in that there was a negotiated political solution to the conflict which led to a major change in government, but without widespread violence and destruction and displacement of people. The land registry and cadastral records were, in general, not compromised. However, there was major structural and spatial reorganisation of state organisations to remove the administrative barriers created under the apartheid system, including major changes to municipal and provincial boundaries, and there were major changes in government agency staff. A challenge that is germane to this discussion was the harmonisation of different land records and filing systems in different agencies. In Edendale, there was violence and displacement of people, but municipal records, land registry records and cadastral survey records were not destroyed.
The components of project management best practice relevant to this discussion include project governance -e.g. the criteria that determine project success and deliverables, how changes such as changes in design, drawings and documents are managed, the procedures to resolve issues that arise, and communication protocols. Added to this are performance and risk management -the identification, monitoring and control of performance (e.g. time, cost, quality) and risk mitigation strategies. Relevant procurement management components are the procurement procedures themselves and then the evaluation of on-site personnel qualifications and stakeholder management involving input and communication with, among others, the people who should benefit from the project (PMI 2013). Earnest (2015) submits that, primarily due to the post-conflict environment, some critical issues that are commonly encountered and should somehow be addressed as best as possible in post-conflict projects include: ineffective project planning and preparation, poor levels of accountability, lack of communication and community participation, lack of skills and institutional capacity, poor procurement practice and the management of project quality.
Analysing the above, if any of the issues that Earnest (2015) mentions are present then there may be high risk of project failure. These issues may arise in many projects, but in post-conflict situations many of them are present concurrently and they are difficult to control. Drawing on the author's experience in consulting projects on land administration strategy in post-conflict Somaliland, a member of a project management team on a large offshore construction project involving a number of contracts in different towns and cities, and as a land surveyor on low cost housing projects in the City of Cape Town in the 1980s, a number of actions may mitigate these risks if the situation allows. In many projects there are clients' representatives who contribute to project governance and monitor performance and quality. In large projects clients' representatives may comprise project management teams having teams on each site. Area experts in these teams draw up lists of performance and quality issues that need to be addressed and monitor these by examining reports and holding regular meetings with each main contractor. Inspectors do independent inspections on site throughout the project. Document management systems are agreed, including procedures when there are changes to drawings that the correct drawings are used to build a structure. An inspector may check on site if the correct drawing is being used to fabricate a particular item. Clients' representatives are likely to assign greater attention to issues that pose a significant risk to project success, and these issues may change rapidly in post-conflict environments. For example, the client's project management team surveyors may conduct independent as-built surveys of structures if the possibility of them being the wrong size poses major risks to the entire project.
Also relevant to the Edendale N case, in a low cost housing project variation on the above, in the author's experience in the City of Cape Town, in-house land surveyors assumed personal responsibility for the correctness of the legal parcel layout (the general plan). They did the legal surveys of party wall boundaries as the foundations were laid if there was a risk that the contractor might not construct the walls according to the layout plan. This was also a check that the contractor and the surveyor were working according to the same plan, as changes to layout plans were often not communicated to the land surveyors themselves.
In synthesis, recognising that many factors cannot be controlled in post-conflict and similar farreaching change scenarios, project governance and general land administration governance can be particularly challenging. They have to be addressed as best as possible. Generally, the higher the risks there are in particular aspects of a project, the greater attention should be assigned to identifying and mitigating these risks. One risk mitigating strategy includes having independent project management teams representing the client(s) who perform independent audits and inspections on critical elements of a project.

Edendale History
Edendale was one of the first areas in southern African where black Africans could own land under individual title registered in a colonial registry (Mkhize 2015). A mission was established at Edendale in 1851 (Meintjies 1988), and the first individual title deeds registered in 1861 (Msimang 1975). The mission area is the core of "old Edendale" (Cowie 2001). Greater Edendale, which is commonly referred to as Edendale or the GEA, falls within the Msunduzi municipality. office in Pietermaritzburg. In addition, planning control vested in different government agencies, a legacy of racially segregated spatial planning policies which pre-dated the apartheid era. As a result, in 1994 the GEA had a convoluted and disjointed land tenure record system, which was both inaccurate and out of date (Cowie 2001). Interviews indicate that problems arising from this have impacted much of the current development in Edendale.
Edendale was one of the centres of a localised civil war in KwaZulu-Natal in the 1980s and early 1990s between Inkatha and the United Democratic Front (UDF). Inkatha was formed in 1975.
Initially it had strong ties to the ANC, but rifts soon occurred. The UDF emerged as an opposition movement in 1983, and it became a front for the ANC (Mkhise 2015). The UDF accused Chief Mangosuthu Buthelezi, the Inkatha leader, of collaborating with the apartheid regime (Mkhise 2015). Violent clashes followed soon thereafter, and there is compelling evidence that the apartheid security forces provided logistical support to Inkatha in these clashes (Kentridge 1990, TRC 1998. The worst of the violence occurred in the seven days war in 1990 when an Inkatha group attacked people in non-Inkatha areas in the GEA. Two hundred people were killed, between 20,000 and 30,000 people were displaced, and 3000 homes were destroyed. Many more were killed and displaced in smaller scale conflicts and assassinations before and after that (TRC 1998). Displaced people moved onto vacant land in the GEA, much of it privately owned, and this stimulated further land invasions and land grabbing (Msunduzi Municipality 1992).
It was in this aftermath of the violence, and the uncertainty that followed, that state-subsidised housing projects were undertaken in Edendale.

Method
The author observed the regularisation processes in an Edendale N Wirewall site office as residents in the project and people who claimed that they were supposed to have been allocated a house submitted evidence of their interests. One Edendale N resident was interviewed. The situation was delicate and it was not possible to conduct door-to-door interviews. Sixteen people who had worked on the Wirewall project, or were working on it, were interviewed augmented by further interviews with 25 people outside of the project who provided context relating to the conflict, land tenure administration and land governance in the GEA. Documentary evidence included newspaper articles, archive documents, land title information drawn from the deeds office, land administration data from a GIS for Edendale, surveyor general's plans, and reports, plans and legal documents obtained from land professionals and officials. Some files held by private firms could not be accessed.

State-Subsidised Housing and Msunduzi Land Governance
Noting that the statistics are contested, the South African government claims to have delivered more than 4.3 million housing opportunities to the poor since 1994 (South African Government: Housing 2017). Whereas there are numerous successful state-subsidised housing projects, there have been a number of problem housing projects across the country. In the problem cases, many of the houses have construction defects, which can be attributed to corruption in the awarding of tenders, maladministration, incompetent contractors, and project governance problems.
Municipalities often lacked skilled personnel and the budget for infrastructure to complete projects successfully (Charlton 2013, Tomlinson 2015. The above should be seen in the context of national level corruption and abuse of power that are to be expected in a post-conflict situation, albeit not many would have predicted this in the first years of ANC rule. The recent national experience as reported in books, public protector reports, the daily press and supported by the author's fieldwork indicate that in the face of pressure to support corruption, dedicated civil servants may face intimidation and fabricated disciplinary charges if they seek to do their work diligently. Moreover, members of civil society who have attempted to uncover corruption have been subjected to death threats, intimidation and false arrest (Paauw 2017).
In Msunduzi municipality, corruption, patronage and abuse of power have for long been a problem among local politicians and senior officials, and the municipality fails to maintain good accounting records (Eprecht 2016, Magubane 2018. Political assassinations, frequently involving competing factions within the ruling ANC party, have plagued KwaZulu-Natal in recent years, and Edendale has not escaped them (du Plessis 2017, Harper 2017. Newspaper articles and the author's interviews reveal a history of patronage, factional conflicts, corruption and maladministration in the GEA and other areas of KwaZulu-Natal. Interviewees claimed that state-subsidised housing construction contracts in Edendale were awarded to companies that were not going concerns and lacked the track record to complete a housing project. Companies were formed to complete a particular housing project and then folded, and a new corporate entity created to tender for projects that followed. Thus, contractors could not be held to account for poor quality work after the project had been completed as the company no longer existed. Many companies were not competent to do the projects that they were awarded. Competency, a group of participants argued, was not afforded the necessary decisive weighting in the tender scoring system. At one stage, there had been stand offs between provincial housing officials and municipal officials over project management and control of housing developments. Patronage and political pressure was a general problem. Ward councillors allegedly pressurised officials to have "their people" awarded the contracts. When this occurred, officials who inspected the work were reluctant to hold these contractors to account in fear of the councillor. In addition, allegations of ward councillors encouraging people to invade land and ward councillors grabbing privately owned land by allocating it to "friends" emerged from a number of sources. Lastly, both officials and residents were fearful of some well-connected, powerful local politicians. While Msunduzi municipality has an admirable whistle blower system in place to fight corruption, interviewees did not trust that their identity would be protected if they filed an allegation against a councillor (Int #104, 115, 121, 123, 127, 131, 132).

Edendale N Wirewall Project
It is in the above environment that the Edendale N Wirewall project was implemented and, as the discussion on post-conflict situations and Msunduzi governance above indicates, project governance and land tenure administration problems were to be expected. There were four construction phases in the Wirewall project, which included a total of 705 houses (Greene Land 2014).

Tenure Problems
The first operations management problems started with the administration of changes to the general plans, i.e. the legal parcel layouts, which cover the Wirewall project. The Surveyor General approved the original general plans for the four phases in 1996. The layout was changed and amending general plans were surveyed and approved. The amended plans cancelled some of the parcels on the 1996 plans and created new ones. However, it appears that the original 1996 plans were not withdrawn nor cancelled or endorsed to reflect these changes. Consequently, transfers of ownership of parcels were registered which did not match the parcel numbers and the layouts which were used to build the houses (Greene Land 2014, 2015a).
In Phase 1 and Phase 4 there were no mismatches between the surveyed parcel layouts and what was registered. Registration was done according to the amended general plans and the surveyed boundaries. There were mismatches in Phase 2, however. The Phase 2 general plan was amended in 2000. The amended plan created new parcels in a different layout to the original 1996 general plan and cancelled the old parcels and their parcel identifiers (i.e. the erf numbers). In 2005, the Msunduzi Municipality registered all but one of the parcels in its own name using the original 1996 general plan. The houses, however, had been constructed according to the boundaries laid out in terms of the 2000 amended general plan. Thus, what was registered does not match what is on the ground. The likely solution is to obtain a court order to cancel the original registrations and reregister the properties in terms of the amended general plan. In Phase 3 an amending general plan was approved in 1997 and then withdrawn. However, the houses were built according to the boundaries surveyed in terms of this amended general plan. The Msunduzi Municipality registered all the parcels in its name in 2006 using the original 1996 general plan. A lengthy administrative process can rectify the situation in Phase 3. All the roads and public places will have to be closed.
Then all the parcels will have to be consolidated, re-subdivided and then resurveyed according to what is currently the situation on the ground, and re-registered (Greene Land 2014, 2015a).
Phase 4 poses a different set of challenges. It is an interesting case of family titles being created because the housing subsidy of R16, 000 that was available when the project was conceived was insufficient to build a good quality house. Four subsidies were pooled and four families from the same extended family who qualified for a subsidy were allocated a larger family house. Each family has a 25% share in the land. The families share facilities such as the kitchen and bathrooms (Int #104, 106, 121).
There have been a number of problems with the Phase 4 family tenure system, however. One anecdote was that in some houses when one of the four family members died, the stronger families refused to allow the heir to move in, or they evicted the deceased's remaining family members.
Families have also wanted to separate their share of the undivided shares and split the house into four separate parcels. It is difficult to do this under the current land use planning and regulatory framework, but the feasibility of the option is being examined. A further problem is that although families had to demonstrate some blood connection, some of the shareholding families were not related. Apparently, people who were not related put their names down to share the same house.
The problem was four families were needed as a condition to get the house, and so if an extended family did not have four families to share the house, they sought "additional family" to make up the numbers. However, in some cases the additional family was never given occupation of the house (Int #104,105,121). Supporting this narrative, participant #150 claimed that they had never occupied their family house. They had never been informed that they had been allocated the house and they had never met the people who were in it, and they were too scared to approach the occupants to clarify the situation.
Fraud appears to have occurred in the allocation of some 200 properties in the Edendale N housing project. The ward councillor at the time the houses were allocated allegedly allocated sites to deceased people and to people from remote rural areas in the Eastern Cape who were unaware of the housing project. He then rented out the houses for his own account and sold some of these houses (presumably off-register) (Greene Land 2014, 2015a. It is not clear how the situation arose. The housing subsidy system is predicated on beneficiaries having a valid national identity document, an "ID", providing their IDs to officials, and proving their identity for subsidies to be assigned and registration to proceed.
There are also issues of transfers involving both legal and illegal original owners who have sold their houses off-register. The nature of these transactions has to be unravelled and the manner in which the current occupants came to live in the house needs to be established. One option is to evict irregular occupants and cancel the transfers which are found to be fraudulent. However, people who purchased these houses from the councillor may have done so in good faith and they have lived in the houses for a long time. The other option is to regularise the situation and allocate the house to them or their heirs (Greene Land 2014, 2015a. A question is if the law allows buyers to take ownership of a subsidised house if they do not qualify for a subsidy.

Housing Construction Problems
A second set of problems arose because a number of the original houses were wirewall houses.
The walls were constructed by spraying concrete onto a reinforcing wire mesh. Wirewall houses were first built for flood relief victims in the Msunduzi municipality. They were then used in Edendale N as a low cost housing option. The first problem was that some residents suffered electric shocks when electric wiring touched the reinforcing mesh. The second problem was that after some time the spray-on concrete crumbled off the wire mesh (Int #104, 121).
The provincial Department of Human Settlements appointed contractors to demolish and rebuild the houses. The province handled these contracts and not the Msunduzi Municipality (Int #104, 106, 121). The new houses are reportedly well constructed. However, it appears the builder did not check the boundary beacon locations when they constructed the houses as 371 of the 705 wirewall houses either straddle the legal boundaries or they encroach over building lines. Many of these new houses can be rescued by subdividing and consolidating some of the parcels and by relaxing the zoning and building regulations. However, an initial estimate indicates that at least nine houses will still have to be demolished as they cannot be saved by relaxing the regulations and reconfiguring the parcel and the surrounding parcels as they contravene fire regulations (Greene Land 2014, Int #121). Again, resolving this situation will be a long and expensive process.
A further problem arose where the occupants were moved into temporary accommodation when the wirewall houses were demolished and new houses were built. When the time came to move them back into their new home, some of the homeowners were moved into houses on different parcels to the ones that they owned. In the interim they have improved their current houses and are reluctant to move again (Greene Land 2014Land , 2015aLand , 2015b. One option to rectify the encroachment problem is for the registered owners to sign contracts (deeds of sale) to reconfigure the parcels through subdivisions and consolidations and to transfer the land between neighbours. An alternative is for the municipality to expropriate all of the affected land, consolidate it, and then re-subdivide it to create a situation where each house falls within the boundaries of a single parcel and it does not encroach over building lines. The new reconfigured parcels can then be transferred to the legitimate owners.

Resolving Ownership Problems
A problem in resolving the subdivision and consolidation strategy to clean up the boundary problems is that deeds of sale have to be signed by the registered owners of each parcel to be subdivided. Likewise, expropriation notices have to be served on registered owners, if that strategy is adopted. The problem is that a significant number of the registered owners cannot be traced, or the original transfers were irregular. There are ghost titles due to the fraud mentioned above, dead man's titles where the original owner has died and the parcel has not been transferred to the heirs, and off-register transactions where the current "owner" is not the registered owner. Thus, the titles will have to be regularised and cleaned up before subdivision and consolidation of parcels can proceed through expropriation or through getting the owners to sign deeds of sale (Greene Land 2014, 2015b).
As part of the regularisation process to clean up the titles, in Phases 1, 2 and 3 the results of door-to-door occupancy surveys must be analysed and the most likely owner of the existing parcels identified by verifying transaction documents, oral testimony of the claimants, their family and neighbours and by verifying identity documents. In addition, strategies will have to be devised to deal with cases where the de facto owners might be disqualified from owning the houses due to housing subsidy conditions. This was initiated on site in a project administered by the municipality, but it ground to a halt in 2016 due to contractual issues (Int #140).
Some of the subdivisions and consolidations may be negotiated if the titles of all the relevant adjoining properties are clean. For those that are not, they will most likely have to be decided by a Land Titles Adjustment commissioner in terms of the Land Titles Adjustment Act 111 of 1993.
Fortunately, there is a longstanding land title adjustment programme in Edendale, as it is no simple task to get a programme up and running in a particular area. Moreover, Land Titles Adjustment is a long and expensive process.
To regularise the Phase 4 situation, a door-to-door occupancy survey was in process. The process was stalled by criminal elements allegedly attempting to take control of the community structures involved in the regularisation process. Field workers indicated that they had been subjected to death threats. Once the social survey has been completed and validated, options include letting some of the occupants continue to live in the house and to move others into a house elsewhere and splitting the home into four individually owned units if it is feasible to alter the construction and amend the zoning regulations to effect this (Int #105,121,140).
In summary, due to what appear to be operations management and communications problems in the Surveyor General's office and the Msunduzi municipality, there was a mismatch between what was registered and what was laid out on the ground in Phases 2 and 3. The experiment with wirewall houses was not a success, and exacerbating the situation many of the replacement houses encroach over the legal boundaries. There is compelling evidence of fraud in the original housing allocations, and many of the occupants appear to be in their houses illegally. In the Phase 4 family houses some beneficiaries appear to have been refused access by more powerful family groups, there have been allegations of heirs being refused access to a house, and family members are unable to sell their shares if they wish to move out of the house.
Efforts are underway to rectify the situation and create clean titles. The Msunduzi municipality does allocate funding to the wirewall project in the ongoing revisions of the city's Integrated Development Plan (e.g. Pietermaritzburg-Msunduzi 2011. However, fixing the multiple legal and construction problems will be a protracted, expensive process.

Analysis and Conclusions
In 1994, given the relatively peaceful and orderly transformation, arguably South Africa was not expected to demonstrate many of the attributes of a post-conflict situation that impact land administration and development projects, but it evolved into this situation due to corruption and patronage at the highest level. In Edendale, however, there was a stronger likelihood of postconflict project governance problems occurring due to nature of the local conflict.
Post-conflict development carries special challenges in that quick responses are imperative to creating social and political stability. Innovative strategies that address an immediate problem quickly, such as the group tenure system and the wirewall houses in Edendale N, may be critical.
However, if there is a risk to the intended beneficiaries' tenure security then innovations should be based on a robust body of evidence and not constitute a grand experiment. It is not clear if the spray-on concrete wall design and the particular method that was used in their construction was well established. Phase 4 was an ambitious communal tenure project that appears to have been driven by financial considerations rather than on local and international evidence of all the critical factors that should be present for this to be a success. Implementation should have taken place according to best practice relating to inter alia: the inheritance strategies that beneficiaries are most likely to adopt, disposal of the shares to non-family members, contributions for the upkeep of the structure, dispute resolution in the event of conflict within and between families in a household, methods of enforcing the shared tenure system rules, and practical ways the rights of vulnerable members in a shared house to assert their rights.
A second issue is related to project management and land administration operations management. The discussion in Section 2 indicates that many of the problems described are to be expected in situations such as Edendale. There is a strong argument for "client" representation of different national, provincial and local government agencies in a project management team in a large national housing programme. Agencies might include, for example, the national Auditor General, the province as the subsidy administrator, and the municipality as the project implementer.
Frequent, publicly available, inspection reports should improve transparency and accountability.
Checks might include inspections of records of housing allocations, inspections of legal survey plans (i.e. general plans) against what is intended to be registered and housing layout plans that are used by housing contractors, procurement practices, checks that people who do the work on site are suitably qualified before they are allowed on site, independent audit surveys of boundary positions and building foundation positions, and checks that the parcels delivered to beneficiaries are the same ones that are being registered. Likewise, the operations management procedures in land administration agencies might be subject to external inspections when major change occurs. If there is the political will at the highest level the more likely the above will be occur.
To conclude, South Africa is distinctive in that it did not experience the major turmoil that occurs in many post-conflict situations, and most development projects were not undertaken by international agencies. Edendale, however, did experience localised violent conflict, and there is compelling evidence of violence, intimidation and corruption. Although the land administration organisations were not severely impacted structurally, inefficiencies can be expected in land administration operations and in development projects. Part of the project governance challenge is to address both tyranny and inefficiency while projects are underway. An important strategy involves both micro and macro level inspections and audits by clients' representatives who should be external to local government organisations.

Acknowledgement
The study was funded by the Canadian Social Science and Humanities Research Council project 435-2014-0399 Land Transactions in South African Cities.