Renewed emphasis in policy discourses on systematic land titling to solve insecure tenure in South Africa is understandable. A staggering two thirds of the citizenry hold off-register land rights. Converting these rights to title deeds may seem self-evident, but our research reveals major stumbling blocks. The article suggests that for a system of land records to succeed, its design must take into account well understood and familiar local and customary processes for holding, using and transmitting land in urban and rural areas.

Introduction: the issues

Many issues cloud the mattter of secure and unequal land tenure in South Africa. The role of title deeds is a particularly thorny one. In the mid-1990s a range of protective land-tenure legislation, applicable to unregistered land rights across the rural and urban landscape, was passed. This was in response to the urgent need to address tenure insecurity in the aftermath of the transition to democracy, in line with the Constitutional injunction (s 25) for secure tenure for all South Africans.

These laws irrevocably redefined land-tenure relations in South Africa. The poor majority on farms, in informal settlements and in communal areas could claim certain 'rights' to land without requiring authentication by formal title. These nevertheless had to be balanced against other rights within the legal property structure. The expectation was that there would be a more equitable spread of property rights and a shift in power relations, thus effecting a change in the socio-economic structure of society.

Two decades into democracy, there is little evidence that a significant shift in the formal property structure has occurred. In local communities, there is a vibrant process of asserting claims and effecting transfers, but at the level of central government, land rights, though more protected, are legally murkier than before. While the 'rights' laws have significantly mitigated the worst effects of arbitrary evictions and other forms of dispossession, they have proved to be a short-term palliative rather than a durable long-term solution. Since their passage in the mid-nineties, it has been more usual for the courts rather than the executive to affirm the rights in law. The government, far from being the custodian and implementer of these rights, has frequently been the opposing party in litigation. It is no surprise to field researchers anymore to hear community members proclaim that the only people they are prepared to talk to are lawyers in support of potential court action.

In short, there remains an entrenched hierarchy of property rights, with customary and so-called informal rights constantly being relegated to the bare minimum of pre-emptive protection against eviction or land grabbing by the elite. These 'off-register' land rights have not been translatable into legally recognised property rights that can be used to leverage economic benefits. The tension between the old structure of property law and the new rights-based law is evident in the high number of infringements of the new legal standards, e.g. evictions on farms and in urban areas, as well as appropriation of rural rights by traditional authorities and business interests, particularly in the mining sector. The corollary is the continued reliance on the courts to resolve this tension, rather than having land policies that are supported by strong institutions.

This article, which is partly based on Kingwill (2017c), suggests some of the key underlying structural weaknesses that exacerbate the continued legal uncertainties around off-register land rights. These deficiencies are reflected in the following questions, or puzzles:

• Why are land rights in the former Bantustans still statutorily undefined and uncertain? Why are these rights still subject to unofficial appropriations by those with powerful corporate and/or traditional-authority interests? And why does the resulting pushback from local landholders result in endless cycles of conflict, contestation and stalled development?
• Why is the state executive so reluctant or incapable of enforcing rights? Why is the legislature so limited and unimaginative in identifying the underlying causes of system's failure and reforming it?

These questions point to ongoing ambiguities and the inability of government to address them systematically. While many are rooted in the political structures of society, the article suggests some alternative mechanisms to strengthen institutions that have a bearing on securing land rights for all South Africans.

The diagnosis: total systems failure in land administration

Former apartheid-era land administration institutions were dismembered as a result of the restructuring of government institutions following the end of apartheid rule. Following the enactment of laws strengthening individual and collective land rights, from 2000 the focus shifted to empowering traditional leaders in the 'communal' areas of the former homelands; in contrast, in urban contexts the focus shifted towards issuing title deeds.

Neither of these approaches resulted in rebuilding a land administration system for the country as a whole. Post-apartheid policy and law makers have in general not tackled the urgent task of rebuilding and remodelling an appropriate administrative infrastructure – one that is able to adjudicate, hold and transfer all rights in the country, including off-register customary and 'informal' rights. An effective land-administration infrastructure should be capable of allowing, and fairly and consistently dealing with, all types of land rights in both rural and urban areas.

Case material from the former Transkei suggests that land rights are generally secure on the ground (i.e. locally). The prevailing norms governing their tenure are well understood at familial and community levels. However, in the world at large these rights are not so secure. When land developments arise, state-regulated administrative processes at higher levels of authority (from municipal through to national levels) are needed to balance local property rights with those of third parties who wish to invest financial and/or political capital. These may include the government’s investing in infrastructure, or local property-rights holders’ wishing to invest in productive enterprises. These processes are, however, cloudy and inconsistent, or at times entirely absent. The authority, controls, social capital and patronage networks of traditional elites operate on the outside of a clear system of land management, as do the claims of real or potential investors in the agrarian and commercial economy.

As a result, land procedures are mired in obfuscation, competing spatial and administrative jurisdictions and sources of authority – and they lack legal definition. In cases of land development, the lack of clarity results in lengthy, unresolvable processes of adjudication, negotiation and ongoing contestation.

Greater administrative accountability is needed to protect local rights and encourage appropriate social and economic development. The Interim Protection of Informal Land Rights Act, intended as a temporary piece of legislation, is precariously poised. It provides only a thin veil of protection for ordinary people's land rights – usually held in inter-generational family structures – against the predations of traditional elites and commercial interests when land developments, including mining and tourism, are proposed. Without institutions to adjudicate the respective claims, contestation results – which discourages investment. Rights are routinely eroded by a powerful evocation of the supposedly customary idea of the 'collective interest' that, in the absence of robust institutions, invariably benefits the powerful. Economic and spatial inequality are perpetuated.

Likewise, in towns and cities, where informal settlements are burgeoning, local government institutions have no land-administration systems to manage local land records; hence the claims of urban informal rights holders are subject to ongoing contestation. Such claims relate to contestation between individuals within families and communities, as well as to national and local government institutions responsible for infrastructure and service delivery. In addition, rights holders frequently need a 'verified address' (or residential tenure certificate) to obtain credit (including lay-byes) or subscriptions such as cell phones, or even to vote in government elections.

Comprehensive title-deed registration is not the solution

It is entirely fair that policy makers, as well as many beneficiaries of land reform, regard the extensive issuing of title deeds as the best form of redress for past and continuing disparities in tenure rights. The South African system of deeds registration, with its highly developed administration system, is regarded as one of the most rigorous in the world. It is no wonder that the fantasy is to extend it to all South Africans in order to solve the problem of a discriminatory land tenure dispensation.

In reality, the Deeds Registry has not proved to have the necessary capability to meet the tenure needs of the various rights holders. A significant body of research, including my own (Kingwill 2014 a), has revealed that this system employs legal norms that do not correspond with the social norms that the vast majority of off-register rights holders draw on in their daily lives. What those who wish to convert unregistered rights into title deeds ignore, is that both the land and the land owner need to be 'eligible' for registration in the Deeds Registry, since there are strict rules governing 'registerability of land'.

Registerability requires one-to-one relationships between a linearly defined property object and a single legally identifiable property holder, whether an individual, a small or large group, a trust or a corporation. Whatever the composition, it must be a 'legal person'. These 'linear' property relationships do not, however, correspond with the de facto relationships that regulate tenure in most African families and communities. Customary property rights are constituted in a matrix of social relationships that regulate access to, and control of, land and other resources. These rights are centred in extended family relationships that ensure succession of property over generations, and they spread outwards to include various levels of community land uses, authority and leverage. Many of these 'layered' relationships persist even after the title deeds have been issued (Kingwill 2014 a).

The requirements of registerability (as currently conceptualised in law) therefore automatically exclude most socially regulated tenures in urban and rural areas. While the ‘official’ process should, in theory, supersede the widespread use of local institutions – which draw on locally understood norms, protocols, procedures and witnessing mechanisms to effect transfers – evidence shows that these practices continue to regulate tenures on the ground.

The assumption that registration will 'transform' rights has, therefore, proved to be delusionary and misleading. This is particularly so where the arguments for registration assume that these rights will be transformed into property that could be utilised or exploited as an economic asset. This assumed process of transformation should be relegated to the art of magic, or at the very least an untested ideological assumption.

Instead, in a recent publication (Hornby, Kingwill, Royston & Cousins 2017), we have drawn on a more empirical approach, using field work to measure why the approach used in the Deeds Office continues to clash with the normative approaches and practices that most people with off-register rights use in their daily lives.

Towards alternative proposals

I have argued that, given their strong social and local characteristics, socially regulated forms of property are difficult to absorb into a nationally regulated system of property rights. They are thus forced into legal obscurity through a lack of adequate legal recognition and administrative support. Empirical research has shown that attempting to move them into the Deeds Registry, with its tight regulation of survey measurement, boundaries and identified owners, renders them out of sync with the way in which they are managed in reality. As a consequence, the vast majority of title holders who subscribe to the social norms described above, do not transfer subsequent transactions in the Deeds Registry. The properties remain in family ownership, but the name of the registered owner on the title deed is not that of the current owner(s), but usually that of an ancestor (in some cases a seller), which technically invalidates the ownership by the current owners.

Designing alternative forms of property recognition would be a complex process, but this is no reason to avoid it. The alternative approach which we advocate is to embed land rights in a set of well-aligned land-management and governance institutions to protect land rights at the level of family holdings, while following a clear set of guidelines as to how to adjudicate these forms of evidence. Strengthening family or individual rights would bolster the claims of rights holders against the claims of others, such as developers, traditional authorities and government, and ensure that land developments present opportunities rather than threats.

To sum up: to make these socially regulated, off-register rights real and to confer the legal recognition as 'ownership' require the development of strong legal institutions to recognise, hold, arbitrate, adjudicate and defend rights in a form that is cognisant with (a) socially recognised practices in families and communities, managed at local levels, but adjudicated in terms of national standards; (b) held in local databases that are designed to accept new forms of evidence, but set up in such a manner that they are managed locally but regulated and co-ordinated by the Deeds Registry to ensure an integrative system; and (c) the norms and principles enshrined in the Constitution regulating gender equity, administrative justice and human and cultural rights.

Conclusion

Policy makers continue to resist remodelling land-administration institutions in a way that aligns them with de facto local and socially regulated institutions. We argue that this is a dangerous practice, since the neglect of property rights, even where they may be relatively securely held locally, seriously curbs social and economic development. The fact that property is not registered does not mean that property relationships are absent. They are constituted and secured within families and communities. To be even more beneficial, these relationships need to be linked to state institutions more effectively, but always in a way that acknowledges locally legitimate social processes, which will endure whether or not title deeds are issued.

The key is to understand what 'is' and enhance the robust processes that already exist, whilst bringing them in line with national laws, procedures and constitutional principles. We believe that, with sufficient political will, this proposal is possible and do-able, albeit over time and in the face of difficult challenges.

Written by Rosalie Kingwill, Research Associate, Institute for Poverty, Land and Agrarian Studies, University of the Western Cape. This article first appeared on Econ3x3.