The consequences of restricting rights to land: understanding the impact of state-land tenure policies in Sri Lanka

Abstract Land tenure can be defined as the mode by which real property is held or owned, or the set of relationships among people concerning land. Understanding land-tenure systems has been key to achieving sustainable development because the terms of these arrangements structure the connection that people have with their lands, influencing what they do with it and how they treat it. Using a two-stage mixed-method design based on scenarios drawn from Sri Lanka, this article demonstrates that affecting control over land tenure and land rights do not always lead to predictable outcomes. Policy interventions often lead to the creation of new and unintended categories of relations between people and the land that run counter to original intentions. The research also demonstrates that relationships to land cannot be adequately captured by looking at legal rights, but rather requires identifying how people perceive or interpret them.


Introduction
Land tenure can be defined as "the mode by which land is held or owned, or the set of relationships among people concerning land or its product" (Hickey and Mitlin 2009; see also Durand-Lasserve and Royston 2002;Payne 2002). Systems of land tenure are often simplified and dichotomized in terms of publicly (or state) vs. privately owned lands. However, most land-tenure systems in the world defy such easy categorization, containing elements of both groupings and existing on a continuum (Lambin et al. 2014;Hausner, Brown, and Laegreid 2015;Boserup 2017).
Understanding the intricacies presented by landtenure systems has been a key feature of sustainable development because the terms of access and/or ownership essentially mediate or structure the relationship that people have with their land, influencing what they do with it and how they treat it (Antwi-Agyei, Dougill, and Stringer 2015;Felson 2017;Behnke 2018). As such, there is a growing body of literature devoted to understanding how specific land-tenure systems operate empirically and to exploring the social, economic, and ecological structures/effects that these land-tenure systems produce. This body of work assumes that obtaining enough knowledge about these associations may enable policy makers to influence the relationships people have with their land in a productive way to achieve optimum use of land resources (Lambin et al. 2014;Antwi-Agyei, Dougill, and Stringer 2015;Seghezzo et al. 2017).
However, findings of multiple studies on this point demonstrate that initiatives to control land (irrespective of whether the object of control is privatization or nationalization of land) by the state or a state-like authority often result in a number of unintended consequences. For example, this was the case when twenty out of the forty countries in sub-Saharan Africa nationalized all lands and extinguished private freehold ownership between 1975 and 2000 or when the Urban Land Ceiling and Regulation Act (1976) of India sought to bring large vacant private landholdings under public ownership or control. In both instances, the result of these policies was that a substantial proportion of the urban population was forced into unauthorized settlements and encroachmentsironically the very types of communities that these policies were intended to prevent (Payne 2002). As such, what seems clear is that any attempt to control the terms of land tenure cannot be implemented on a heavy-handed basis. These initiatives need to both qualitatively understand the subtle varieties of land tenure that exist in any given sociopolitical setting (UN-HABITAT 2012; Barry and Asiedu 2017) and the attitudes and perceptions of the people involved (Barry and Roux 2014). This latter point in particular has not received much serious attention.
This article makes several contributions to current debates in the literature on land tenure and policy controls by undertaking a focused, qualitative case study on state land-tenure policies in Sri Lanka and examining how they affect the relationship of people to the land. This investigation reaches several conclusions regarding the need to understand the rich complexities of land rights on the ground prior to attempting land control. First, it is established that any attempt to restrict rights to land by controlling land tenure results in the creation of ever more complex sets of relations between people and land, accompanied by a cascade of unintended social and environmental effects. Second, the study argues that gaining de facto understanding of complex land-tenure systems as they exist in practice should also be complemented with an understanding of the ways people perceive or interpret their land rights under the tenure system. Related to this point is realization that there are no absolute standards by which the security of land tenure or rights over land can be defined. For encroachers who have lived in a settlement for many years, their perceived security of tenure and rights over land may be indistinguishable in practice from households living in legal housing.
The next section highlights the design of the study and the third section documents its main findings. I then discuss the significance of this work in terms of the wider issues pertaining to land tenure. The final section considers the policy implications of this study, specifically in the context of South Asia and sub-Saharan Africa.

Study design
The study design used in this research contains two stages of investigation. In the first stage, I pursued several independent lines of evidence in attempting to understand how the land-tenure system works in the state lands (lands that belong to the state and have been allocated to the populace subject to certain conditions of residence and cultivation) of Sri Lanka. This pursuit involved a documentary analysis of a number of statutes and regulatory instruments that address the governance of state lands. The investigation was supplemented with key-informant interviews (n ¼ 20) with senior policy advisors from the country's Ministry of Lands and Parliamentary Reforms to understand policy decisions undertaken by the government over the years to control landtenure systems in state lands.
In the second stage, I carried out exploratory fieldwork in two adjacent farming villages (comprising a total of 186 households: Village A ¼ 96 households and Village B ¼ 90 households) in North-Central Province of Sri Lanka. The reason for selecting these particular communities was twofold. First, the settlers of the villages have been occupying state lands such that their relationships to and rights over the land could legally be mediated by the government. Second, the government attemptedas part of a larger agricultural development project carried out in the provinceto exert systematic control over land-tenure patterns in these villages (by limiting residents' rights to the land in terms of restricting subdivisions, selling or mortgaging, and subletting) over the course of several decades to achieve higher agricultural yields.
When carrying out data collection in this second stage, I first obtained village-level population statistics and household data, as well as existing data on land-use patterns, from the offices of the divisional secretariat and constructed several preliminary categories of tenants based on their relationship to the land. I created three initial categories of tenure: qualified freehold owners of state land (i.e., the holder was granted ownership rights to the land subject to certain restrictions), qualified leaseholders or renters, and illegal encroachers. Second, I engaged in ethnographic work over the course of six months to investigate the de facto land-use practices of the residents in the sample villages. Specifically, I utilized participatory observation (with the observations recorded in field diaries) which enabled me to gain a more comprehensive understanding of the nonofficial or 'shadow' categories of tenancy arrangements and actual land-use practices by the villagers. The findings from these observations further complicated the simple three-fold categorization that I had initially developed. Third (and as a logical extension of participant observation), I conducted a combination of semi-structured and unstructured interviews in each village (n ¼ 30 Â 2) and focusgroup discussions (n ¼ 1 Â 2) with a purposively identified mixed group of tenants spanning the full continuum of tenancy categories (both official and unofficial), as previously identified. These interviews were carried out to gain insight into the perceptions of the tenants regarding the legality of their residence and their rights to the land.
Finally, I conducted a household survey in one of the sample villages (Village A) to quantitatively measure the tenure security of residents and (in the case of unlawful residents such as encroachers) the risks of continuing residence as well as their openness to leaving. The survey consisted of a short questionnaire that was administered in person to the head householder in each of the cases. I focused on two main questions. In Question 1, the householders were asked to rate (via a self-rated Likert scale) their perceived security of tenure (1 ¼ little to no tenure security perceived/5 ¼ high degree of tenure security perceived). For Question 2, I asked householders to rate their willingness to leave or relocate (1 ¼ high willingness to relocate/ 5 ¼ low willingness to relocate).

State-land tenure policies in Sri Lanka
An examination of a variety of statutes and regulations pertaining to state lands in Sri Lanka helps to provide a comprehensive picture of how land tenure works in the country. After the British colonized Ceylon (as it was known at the time) in 1798, a substantial amount of land was acquired by the Crown through the Crown Lands Encroachment Ordinance No. 12 of 1840, and this property was subsequently gifted or sold to Europeans or the local aristocracy. Following Sri Lanka's independence in 1948, the Crown lands were reclassified as state lands under the Land Development Ordinance of 1935 and the State Land Ordinance No. 8 of 1947. However, the newly labelled 'state lands' imposed the same rules of ownership and tenure as Crown Lands, save for the fact that the lands were now owned by the Government of Sri Lanka and not the Crown (Paranage 2018).
In the mid-1950s, the Sri Lankan government started allocating state lands to peasants and farmers via land grants as part of a number of large-scale agricultural development projects so that the lands would be used for cultivation. However, the lands that were allocated carried a number of restrictions pertaining to ownership and tenure. These legal constraints enabled the state to mediate the relationship between the grantee and the land. By controlling the system of land tenure, the government sought to manage the land use and cultivation behavior of the settlers so as to maximize agricultural gain. This would not have been as readily possible if the relevant authorities had given the settlers full (freehold) ownership of the land and left them to their own devices.
There are two notable examples of such restrictions. First, the lands are granted to the owner subject to a rule of minimum subdivision, which means that the original receiver of the grant can pass on his rights to the land (by selling it or under the laws of succession) only if this action does not entail dividing up rights to the land. The rationale behind this rule is based on economic grounds (predicated on a long legacy of academic and policy research) maintaining that land fragmentation is contrary to sustainable developments in agriculture (Niroula and Thapa 2005;Hartvigsen 2014;Latruffe and Piet 2014;Teshome et al. 2016). Accordingly, this rule preventing subdivision interferes with the usual practices of succession (for instance if the original land owner had more than one offspring) and makes no manifest provision for any contingencies. Second, grantees are restricted from leasing or renting to nonrelatives. This rule has been established so that the land is not leased to nonfarmers (as only farming families qualify to receive land grants in the first place), and thus, the land would only be used for agricultural cultivation.
Consequences of restricting rights to land: the creation of 'shadow' subclasses Ethnographic fieldwork carried out in the two sample farming villages that were subject to the abovementioned restrictions on minimum subdivisions and leasing revealed that placing restrictions on rights to the land has created several tenure categoriesofficial, unofficial, and illegal. Table 1 identifies the tenure categories that appear in both sample villages with a brief description of each of them.
This wide variety of 'shadow' subcategories has come into existence as an unintended consequence of attempting to restrict rights over land. Restrictions against leasing and subletting have led to unauthorized tenancy arrangements being created through the use of promissory notes. Likewise, restrictions against subdivisions without considering the implications for property succession have caused both regularized and unregularized encroachments and unauthorized subdivisions of land. Fieldwork has revealed that the majority of encroachments occur during the 'second' and 'third' generations of settlersa situation that takes place when the original owner of the land passes his rights to one (often the eldest) son/daughter and the recipient ousts the rest of his or her siblings (without any legal claim) to the property. This growth of 'shadow' tenure categories due to attempts to achieve land control has many social and environmental repercussions. The most obvious instance of this phenomenon is the manifold social and environmental effects created by the nonregularized encroachments. As mentioned above, this survey was conducted in two villages where the principal occupation was irrigated farming. While most of the other tenant classes (i.e., unauthorized subdivisions, promissory noteholders) reside in the zones identified and separated for household building, the nonregularized encroachments often squat in the areas reserved for the safety of the irrigation infrastructure (e.g., tanks, dams, and canals). Further, nonregularized encroachments, when living close to the water-distribution canals, often dispose of their waste into the canals which causes blockages and creates widespread mosquito-breeding spaces. This is an important finding, given that existing literature argues that sustainable land management calls for land resources to be utilized not only to service human needs but also with a view to ensuring the sustainability of ecosystem services (Grumbine 1994;Berkes et al. 2000;Carpenter et al. 2009;De Groot et al. 2010). Thus, it is possiblein this contextto argue that the creation of certain shadow subclasses like encroachments becomes a severe challenge to sustainable land use.
Further, evident tensions exist between the nonregularized encroachments and legal residentsfriction that did not extend to the other shadow-tenure categories. These frictions are a by-product of the ecological concerns mentioned above, namely the environmentally damaging waste-disposal practices of encroachments. The consequences (especially health and sanitation concerns) of harmful waste disposal extend to the entire village community; thus, conflicts and verbal altercations between encroaching groups and other residents are a frequent occurrence.
The practices of waste disposal, and the overall indifference toward local ecological conditions, generally tend to distinguish the nonregularized encroachments from other legal and 'shadow' subclasses of tenants since almost all other subclasses (irrespective of their legality) conform to more acceptable standards of cultivation and waste disposal. This distinction also applies between regularized and nonregularized encroachments, as regularized encroachments have been given land plots by the state in the zones identified for building households and cultivation. It also appears that once nonregularized encroachments are regularized and integrated into the community, they tend to follow existing regulations, customs, and practices regarding cultivation and waste management. It is, however, ambiguous whether actual or perceived tenure security has any correlation with how inhabitants use their lands, although other researchers have investigated this hypothesis on land-tenure patterns (Payne 2002;Hausner, Brown, and Laegreid 2015).

Land rights and 'perceived' land rights
The second important result is the discrepancy between the actual legality of a particular tenancy category or 'shadow' category mentioned above and the perceived 'legality' of residence and tenure security. To highlight this issue, I compared the results of the household survey in which participants of each of the seven subclasses listed above were asked to rank themselves based on: 1) their perceptions regarding the security provided by their tenure arrangement (1 ¼ no security of tenure/5 ¼ very secure in their tenure arrangements) and 2) their willingness to relocate (1 ¼ high willingness to relocate/5 ¼ low willingness to relocate) to the actual legal security provided to members of each subclass. In Figure 1, the horizontal axis represents both the perceived security of landholders and tenants (blue) and their willingness to relocate (red). The vertical axis indicates the actual legal security provided to members of each subclass represented on an ordinal scale (the actual legal security of each class was determined by an analysis of laws and regulatory instruments regarding tenure, land rights, and encroachment as well as via key-informant interviews conducted with senior policy advisors of the state). Category 1 has the lowest actual security of tenure and legality and Category 7 has the highest actual security of tenure and legality.

Encroacher (nonregularized)
This category comprises persons who illegally occupy parts of state lands and may have also constructed housing or other facilities on it. This class of persons is at the highest risk of being evicted.

Tenant in unauthorized subdivision
Tenancy in an unauthorized subdivision can occur if the landowner decides to unofficially (and illegally) subdivide his/her land and lease out part of it to another.

Owner of unauthorized subdivision
The rules against subdivisions prevent landowners from subdividing their lands into smaller parcels. However, in the case of a landowner who has more than one offspring, the land has often been de-facto separated and owned by the siblings, each of whom has constructed homesteads in her or his respective parcels of land.

Tenant on contract (promissory note)
This category comprises tenants who are not relatives of the original owner of the land but to whom the land has been (unofficially) leased. A common practice in the villages is to lease out lands through the use of 'promissory notes' (a signed document containing a written promise between the original owner and the lessee) which serves as a lease contract between the parties. The extent to which these documents are legally binding remain ambiguous.

Encroacher (regularized)
This category comprises persons who previously encroached (illegally occupied and/or constructed premises) on state-lands but have been subsequently regularized by the state. They are now 'legal' in their occupation.

Approved leaseholder/tenant
This category comprises persons to whom the land has been leased by the original owner. These leases are approved by the state, as the lessee is a relative of the original owner of the land.

Qualified freeholder
This category comprises the original freeholders to whom land has been granted by the state. They 'own' the lands subject to restrictions in subdivisions and leasing.
Table 2 and Figure 1 demonstrate that there is no correlation between the actual legality and the perceived tenure security of residents, regardless of tenure category. Generally, the perceived tenure security of the residents was high across all tenure categories. The interviews and focus-group discussions that were carried out as an extension of the ethnographic fieldwork provides some insight to explain this discrepancy.
In relation to Categories 6 and 7 (approved leaseholder/tenant and qualified freeholder, respectively), the reasons for the high level of perceived tenure security is self-explanatory, given their residence is recognized by formal legal institutions. Category 5  (regularized encroachments) also enjoy formal legal recognition. By contrast, Category 4 (tenant on contract through a promissory note) is not recognized in terms of overall legality. However, respondents from this class perceived their tenure as being secure because it is based on a written document that they regard as a contractual obligation owed to them, irrespective of whether such a contract would be enforceable in a court of law. This situation also implies that there is no one standard with which the authority of laws can be judgedmembers of Category 4 clearly perceive their tenure as 'legal' based on the promissory note. For them, the breach of promise is 'inequitable' and therefore 'illegal.' Category 3 (owner of unauthorized subdivision), Category 2 (tenant in an unauthorized subdivision), and Category 1 (nonregularized encroacher) are all essentially encroachments. This outcome occurs because these categories of residents all occupy unauthorized land (whether occupying or leasing an unauthorized subplot of a larger parcel of alienated state land in the case of Categories 2 and 3 or a case of encroachment on a completely new piece of state land in the case of Category 1). These categories are similar to Category 4 in that they are afforded no manifest legal protection (in fact, their residence is a violation of the tenure restrictions imposed by the state). However, unlike Category 4, they cannot rely on a contractual obligationmaking their overall position even more vulnerable than members of Category 4. Notwithstanding their vulnerability on two fronts (i.e., absence of a manifest legal right and a contractual obligation), these categories also appear to have a high degree of perceived tenure security. Three explanations are relevant.
First, encroachments have been established for long periods of time and, for those encroachers who have lived in a settlement for many years, their perceived security of tenure and rights to the land is often indistinguishable in practice from households living in legal housing.
Second, many encroachments in Category 1 and Category 3 are actually the children of the original (first-generation) landowners. As such, interview results demonstrated that these (displaced) offspring feel a strong sense of entitlement to the properties in question and they assert that the state-land parcels rightfully belong to them as well as to their siblings, a contention that consequently feeds their perceptions of tenure security. Further (similar to the case with Category 4), for members of these categories (1 and 3) there appears to be no standard of absolute 'legality' or authority. Rather, they willfully disregard the 'laws' preventing subdivisions as unfair and maintain that their 'right' to the land outweighs any law imposed by the government to prevent a natural line of succession.
Finally, a comparison of actual (legally recognized) tenure systems reveals that perceptions of legality, entitlement, and tenure security are more useful indicators for predicting land use-related behaviors and practices. More specifically, there is an inverse correlation between perceived security of tenure and willingness to relocate. On one hand, when the householders believe that their security of tenure is high (as outlined above) they are not willing to relocate or 'give up' their perceived right to state lands. On the other hand (as in the case of perceived tenure security), there again seems to be no correlation between willingness to relocate and the actual legal circumstances. In particular, the perception of having a sense of 'ownership' over land as rightful property appears to motivate the encroaching families to continue to reside on state lands despite repeated threats of eviction, up until the point at which they are regularized. In fact, 'regularization' for these households represents a concession on the part of the government, a recognition of their rights as the displaced children of the original grantees of land. The eventual regularization of encroachments can also (alternatively) be viewed as the failure of the state to 'control' tenure systems and to restrict the land rights of its residents.

Discussion
It was initially noted that effective and sustainable land use and management are essential to achieving the goals of sustainable development (Smyth and Dumanski 1993;Bindraban et al. 2000). Understanding land-tenure systems becomes very important in this context, given that tenure systems are what mediate or structure the relationship that people have with the land, influencing what they do with it and how they treat it.
The findings that I have presented with respect to Sri Lanka make several important contributions to contemporary understanding of the complexities of tenure systems. First, this study documents that heavy-handed control of land tenure and rights results in many unintended consequences (in the current case, the creation of 'shadow' tenure categories). In accordance with the observations of Payne (2002) who focused on India and sub-Saharan Africa, attempting to control land tenure to elicit a particular objective often seems to result in outcomes that defeat or run counter to original policy objectives. This has also been true in Sri Lanka. For instance, the prior analysis notes that the driving motivation of the government in restricting subdivisions has been to curtail land fragmentation which would have had the effect of reducing agricultural productivity of land resources (Heston and Kumar 1983;Bentley 1987;Sabates-Wheeler 2002). However, in the sample villages, restricting subdivisions has not made any difference to land being de facto fragmented. Individuals dispossessed of land typically strive to find a variety of extra-legal methods to subdivide the land into consistently smaller parcels. Common strategies in Sri Lanka have included using promissory notes, encroaching, or creating unauthorized property subdivisions and tenancy agreements. In other words, attempting to control land use in an effort to prevent land fragmentation is apt to have little effect and to create a number of shadow-tenant classes that continue to fragment and use lands by various extralegal means.
The findings of this work also generally conform with the established currents of contemporary research on sustainable land use and management, particularly the metaphor of a continuum of land rights. For instance, it appears that neither the two major ordinances in Sri Lanka on land acquisition and redistribution (the Crown Lands Encroachment Ordinance No. 12 of 1840 and the Land Development Ordinance of 1935), nor subsequent legislation on the matter, have considered the potentially wide and complex spectrum of tenure arrangements that existed on the ground (see also Paranage 2017). Further, this work also draws attention to the idea that to better appreciate the behavior of the various groups of tenants, it is also necessary to pay attention to the ways in which the tenure system and land rights are interpreted or perceived by the various social groups. In fact, the findings of this study suggest that the members of tenure categories do not distinguish between the rights granted to them by the law of the land and the rights that they believe are owed to them. This pointthat there is no absolute standard by which to separate real rights and perceived rights to the landis demonstrated by the quantitative results. Regardless of the actual legality afforded to them by the laws, the tenants of vulnerable categories still rank the perceived security of their tenure as high. This finding is also important as it firmly establishes the importance of studying tenure systems from the standpoint of the perceptions and interpretations of the residents rather than in isolation.

Conclusion
This article is based on a case study that examines the state land-tenure policies in Sri Lanka and how these policies affect the relationship between residents and the land. The study reinforces the importance of acknowledging the diversity of landtenure and use patterns that may exist in a given sociopolitical context, while also stressing the importance of understanding how people actually perceive their land rights. By highlighting the consequences of ignoring the complex relationships that mediate the rights that people hold to land, this work contributes most directly to the growing body of literature on sustainable land management and evolving land rights. The work reported here is specifically germane to the case studies carried out in sub-Saharan Africa (notably in Malawi, Namibia, Angola, Mozambique, and South Africa) and in Southeast Asia by various international institutions such as the Global Land Tool Network (GLTN) and UN-HABITAT (see UN-HABITAT 2012;UN-HABITAT 2013;Chigbu et al. 2017). By focusing on the importance of identifying the perceptions that people maintain with respect to land, this study also contributes to an underexplored issue pertaining to sustainable land use. The ways in which residents perceive and interpret land rights should also be regarded as vital to gaining a proper understanding of land-use practices and should guide the direction of future research.
In addition, this study is also of concrete importance from a policy-oriented standpoint. For instance, this investigation makes evident that the rule of minimum subdivisions posited by the Sri Lankan government to control land rights has disregarded the usual practices of succession and has created no room for contingencies. It is this inflexibility that serves to create the 'shadow' tenure categories as well as to inform the perceptions of people as being victimized by the state. Quite apart from paying attention to existing tenure categories at the stage in which land policy is drafted, one approach that the government could have taken would have been to exert limited control at inception. This means that policy makers could have enforced the rule of minimal subdivisions while also accommodating other legal arrangements including legal tenancy agreements, promissory notes, and legally enforceable leases. Such an approach would have given members of the community the necessary leeway and flexibility to adapt to a new set of regulations designed to reconfigure land use. Further, instead of considering the 'shadow' subclasses as illegal ex post facto, the government could have acted to integrate these wide-ranging subclasses and tenure categories into the community by providing legitimation and recognition. This objective could have been achieved by legalizing the use of promissory notes, regularizing encroachments, and tenancy agreements.