See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/5217722

Land Tenure and Property Rights: Theory and Implications for Development
Policy

Article in The World Bank Economic Review · January 1991

DOI: 10.1093/wber/5.1.135 · Source: RePEc

CITATIONS READS

586 4,541

2 authors:

David Feeny Gershon Feder

McMaster University Consultative Group on International Agricultural Research
442 PUBLICATIONS 36,858 CITATIONS 112 PUBLICATIONS 18,877 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Gershon Feder on 01 December 2015.

The user has requested enhancement of the downloaded file.

Public Disclosure Authorized

T HIE WO R LD BANK EC ONO MI C REV I E W, VOL. 5, NO . I: 1 3 5 -IS 3

FILECOPY t,
Land Tenureand Property Rights: Theory
and Implicationsfor DevelopmentPolicy
Gershon Feder and David Feeny
Public Disclosure Authorized

This article explores the nature of property rights systems, their evolution, and their
effect on resource allocation. It is argued that certain institutional arrangements for
land rights have evolved in order to reduce uncertainty and increase efficiency in credit
as well as in land markets. Of particular relevance to developing countries, the article
emphasizes the contribution of public sector infrastructure to effective land rights
systems. An appendix to the article presents a formal model analyzing the effects of
security of land rights on land prices, the intensity of cultivation, and the use of credit.
Empirical evidence from Thailand supports several of the propositions derived from the
model.

The system of private property rights in land found in modern Western econ-
omies is the product of centuries of economic, social, political, and legal change
Public Disclosure Authorized

(North 1981). Most economic analyses presume Western-style exclusive, trans-

ferable, alienable, and enforceable private property rights in land. In this case,
the traditional three pillars of economic theory, namely, resource endowments,
technology, and preferences, are sufficient. The fourth pillar, institutions, can be
omitted without seriously distorting the analysis. Yetboth historically and in the
contemporary world, especially in developing countries, the presumption of
exclusive, transferable, alienable, and enforceable rights is frequently inaccurate
and potentially misleading. In such cases the complex nature of institutional
arrangements in general and property rights in particular needs to be described;
the fourth pillar needs to be specified (Arrow 1985; Coase 1960; Feeny 1988a;
Solow 1985).
This article focuses on how property rights in land affect resource allocation
in agriculture in developing countries. A number of topics are considered, in-
cluding conceptual issues concerning the nature of institutions and the definition
Public Disclosure Authorized

of property rights. We consider the impact of land rights systems on incentives,

Gershon Feder is in the Agriculture and Rural Development Department of the World Bank. David
Feeny is in the Department of Economics at McMaster University, Ontario, Canada. The authors
acknowledge the helpful comments of Clive Bell, John Bruce, Karla Hoff, and several anonymous
referees.
) 1991 The InternationalBank for Reconstructionand DevelopmentI THE WORLDBANK.

135
136 THE WORLD BANK ECONOMIC REVIEW, VOL. 5, NO. I

_uncertainty, and the operation of credit markets. Of particular relevance to

developing countries, we emphasize the importance of public sector infrastruc-
ture in making land rights systems effective. A formal model analyzing the
effects of property rights security on land values, the intensity of cultivation, and
the use of credit is developed. The model highlights the strong interactions
between security of property rights and credit markets. Empirical evidence from
Thailand supporting some of the propositions derived from the model is summa-
rized briefly.

I. PROPERTYRIGHTS AS AN INSTITUTION

In order to consider the role of property rights in general and land rights in
particular, it is important to place these rights in the context of the overall
institutional structure of the society and economy. There are three basic catego-
ries of institutions: constitutional order, institutional arrangements, and norma-
tive behavioral codes. The constitutional order refers to the fundamental rules
about how society is organized-the rules for making rules. Institutional ar-
rangements are created within the rules specified by the constitutional order.
These arrangements include laws, regulations, associations, contracts, and the
focus of this article, property rights in land. The third category, normative
behavioral codes, refers to the cultural values which legitimize the arrangements
and constrain behavior. The constitutional order and normative behavioral
codes evolve slowly; institutional arrangements may be more readily modified.
In developing countries that are undergoing evolution in all three categories of
institutions there is the potential for a lack of congruence among the three types
of institutions. Thus, although the formal legal system may provide for alien-
ability, the transfer of land to persons from another clan or ethnic group may
represent a violation of cultural norms. Similarly, although the constitutional
order may make provisions for private property rights and there may formally
be laws establishing such rights, the corresponding registration and enforcement
mechanisms may be largely absent.
Property rights are an important class of institutional arrangement. In gen-
eral, "property as a social institution implies a system of relations between
individuals. . . . it involves rights, duties, powers, privileges, forbearance, etc.,
of certain kinds' 1 Property rights are then a bundle of characteristics: exclu-
sivity, inheritability, transferability, and enforcement mechanisms (Alchian and
Demsetz 1973). Thus property rights define the uses which are legitimately
viewed as exclusive and who has these exclusive rights. Uses of land may include
hunting, passage, gathering, grazing, cultivation, the mining of minerals, the use
of trees, and even the right to destroy the resources. For instance, in medieval
England and contemporary South India, rights to the crop are private whereas

1. Hallowell (1943, p. 119); for a discussion on the historical evolution of the concept of property see
Schlatter (1951).
 Feder and Feeny 137

rights to the stubble after harvesting are communal (Campbell and Godoy 1986;
Wade 1986). Similarly,in many parts of Sub-Saharan Africa land and tree tenure
are separate (Feder and Noronha 1987). Land rights may further specify the
conditions under which various types of transfer of rights may be affected and
the parties to whom such transfers may be made. Rights also have a temporal
dimension. The institutional arrangements include mechanisms for defining and
enforcing property rights; that is, they include both the formal procedures and
the social customs and attitudes concerning the legitimacy and recognition of
those rights (Taylor 1988). Enforcement depends on a constellation of support-
ing arrangements and mechanisms such as courts, police, financial institutions,
the legal profession, land surveys, record-keeping systems, and titling agencies,
in addition to the social legitimacy of property rights in land.
Four Categories of Property Rights in Land
There are four basic categories of property rights in land: none (or open
access), communal property, private property, and state (or crown) property.
Under open access, rights are left unassigned. The lack of any exclusivity implies
the lack of an incentive to conserve, and therefore often results in degradation of
scarce resources. Under communal property, exclusive rights are assigned to a
group of individuals. Under state property, management of the land is under the
authority of the public sector. In private property, an individual is assigned the
rights.2 These four categories are ideal analytical types. If the group holding
exclusive communal rights is large enough, the distinction between communal
property and open access becomes moot. If private property rights are not
viewed as being legitimate or are not enforced adequately, de jure private prop-
erty becomes de facto open access. Nonetheless the simple taxonomy is useful
for describing property rights systems.
All or some of these categories of property rights may exist in a single society
for different tracts of land. Furthermore, because of the multifaceted nature of
property rights in land, the same tract of land can be categorized under more
than one regime. In many societies, some or all land is constitutionally the
property of the state, but exclusive use rights are given to individuals under a
contractual arrangement with the state. If these use rights are transferable with
few limitations, and if the contract is sufficiently long-term (for example, ninety-
nine years), then for most of the contract's duration there is very little difference
between possession of use rights and full property rights.
The changes in economic relations and in power structures that characterize
the development process generate changing needs for property rights and for the

2. The term common property is sometimes used to refer to property that is classified as communal in
the system used here, but it is also used to refer to open access situations. More generally, common
property refers to situations for which exclusion is difficult and utilization involves rivalry. To avoid
confusion, group ownership is therefore labeled as communal, rather than common property rights. For
more on these distinctions, see Berkes and others (1989).
138 THE WORLD BANK ECONOMIC REVIEW, VOL. 5, NO. I

associated institutions to regulate or enforce them. In the early stages of agri-

cultural development, land rights may be split between individuals and the
community. Individuals are assigned use rights (which can be long-term and
inheritable), although the right to sell land or transfer the use right to nonheirs is
retained by the community. Under circumstances where endowments are similar
across households and land is abundant, such arrangements provide incentives
to individuals to exert effort in tilling land and preserving fertility (through
secure and inheritable use rights), yet they minimize social tensions. Social un-
rest may emerge when individuals lose their land rights, especially to nonmem-
bers of the community, creating a landless class. When technology advances,
however, and endowments of labor and other productive assets differ among
households, the lack of transferability of property rights may adversely affect
productivity. Efficiency considerations thus motivate changes both in the consti-
tutional order and institutional arrangements relating to land rights.
Evolution of Land Rights: The Example of Thailand
Private property rights in land have evolved gradually in response to increases
in the scarcity value of land and therefore the benefits to be derived from more
precise and secure land rights. When land was abundant and labor scarce,
property rights in labor (for example, slavery) were often defined with much
greater precision than property rights in land (Engerman 1973). For example, in
land-abundant, labor-scarce, early-nineteenth-century Thailand, slaves rather
than land served as collateral in financial markets. There was a well-developed
legal system to govern transactions in labor commitments. In contrast, the sys-
tem of usufruct land rights was not as extensively developed.3 In theory all land
belonged to the king. In practice individuals were allowed to use the land for
cultivation, sell it, and pass it on to their heirs as long as they paid taxes on the
land and did not leave it fallow for longer than three consecutive years.
With the increasing commercialization of rice production and exporting of
rice, land values in Thailand appreciated, even though for much of the period
the area cultivated per person in fact increased. The agricultural terms of trade
appreciated as international trading opportunities were opened up and transpor-
tation costs declined. The result was a rice export boom which induced a rapid
expansion in the area under cultivation.
As land became more valuable and frontier areas were brought under cultiva-
tion, land disputes became endemic. The Thai government responded with a
series of procedural and administrative changes. A major new law on land rights
was enacted in 1892. Although it provided a more comprehensive framework
and more standardization of procedures for documenting land ownership rights,
the lack of adequate surveys and record-keeping continued to inhibit the precise
documentation of rights; land disputes continued. In 1896 the government re-
sponded by initiating a cadastral survey in an area in which important govern-

3. A detailed discussionof the evolution of land rights in Thailand is providedby Feeny(1982).

 Feder and Feeny 139

ment officials were also landowners, and in 1901 created a formal system of
land titling. Cadastral surveys covering most of the commercialized areas in the
central plain followed. Surveys were, however, not vigorously pursued in most
other regions or in upland areas. Thai legislation continued to evolve. The result
has been a compromise between the traditional practice of allowing citizens to
bring unoccupied forest land under cultivation as private property and the re-
quirements of the land titling system based on cadastral surveys. The compro-
mise provides for several levels in the security and documentation of land rights.
It is embodied in the 1954 legislation which provides the basis for the current
system in Thailand.
In summary, the current system of land rights in Thailand developed in re-
sponse to the increased benefits of defining property rights in land induced by
the commercialization of agriculture and appreciation in the agricultural terms
of trade. Government officials, as landowners, shared in the gains from titling
and were therefore willing to supply the institutional changes being demanded,
especially in those areas in which they owned land. Their motives also reflected
the desire to provide mechanisms to resolve and reduce the incidence of land
disputes.
This evolution is not unique to Thailand. Increasing population density, ap-
preciations in the agricultural terms of trade, and technological change which
made investments in land quality more profitable have enhanced the benefits
from creating more precise private property rights in land.4 The processes which
helped to shape the historical development of land rights in Thailand and in the
West are very salient today in many developing countries. Population pressure
on land resources is common. Many new technologies have increased the returns
to farming. The demand for institutional arrangements to describe and enforce
property rights in land with more precision is thus a common feature of many
developing countries.

II. CONSEQUENCES OF LAND RIGHTS SYSTEMS: INTUITIVE PROPOSITIONS

Systems of property rights in land, by assigning and enforcing the gains and
losses from actions to agents, have a profound effect on incentives and on the
scope of market transactions in land and credit (where land is often used as
collateral). Here we summarize the nature of these effects.
Incentives
Property rights provide agents with the incentives to use land efficiently and to
invest in land conservation and improvement. The establishment and enforce-
ment of these rights are, however, not costless. When land is abundant, the gains
afforded by enhanced property rights may be more than offset by the transaction

4. See Feeny (1988b), Libecap (1986), Roumasset and LaCroix (1988), and Umbeck (1977) for
examples of the process.
140 THE WORLD BANK ECONOMIC REVIEW, VOL. 5, NO. I

cost of providing for the property rights. If land becomes scarce, however, or
changes in technology create new investment opportunities, the forgone gains
become more important and the provision of property rights in land then has the
ability to enhance productivity. Communal rights may represent the best ar-
rangement for situations in which the opportunities to invest in the quality of the
land are limited and the community is small, but because land is sufficiently
scarce it pays to exclude outsiders from using it. In such a situation communal
rights economize on transaction costs. Outsiders are readily detected, and the
entire community has an incentive to enforce their exclusion. The small size of
the community implies that the transaction costs of regulating use among mem-
bers are not prohibitive. It is often observed in larger communities, however,
that mechanisms for imposing restrictions on individuals' land use patterns
which are harmful to the group's interest are deficient, and communal owner-
ship then leads to efficiency losses. Furthermore, when new market oppor-
tunities arise or new technologies provide large benefits from investments, com-
munal rights may no longer provide sufficient incentives.
Asymmetric Information and Uncertainty
Given the effect of land rights systems on incentives, it follows intuitively that
risks to the possession of such rights (for example, the risk of state expropria-
tion, of private challenges to land rights, or of tenure agreement cancellation
faced by a tenant) will hurt production and investment. Here we wish to empha-
size a particular aspect of efficiency loss due to asymmetric information. In the
early stages of agricultural development, transactions in land take place mainly
among members of the same community. Information is thus fairly symmetric:
the identity of those who possess transferable rights over specific tracts of land is
reasonably well known to all members of the community. With more advanced
stages of development and increased mobility of individuals and entrepreneurs,
transactions among individuals who are not members of the same community
are more frequent. As a result, the scope for asymmetric information, and hence
land disputes, increases. The price of land will then not reflect its true social
value, and the extent of land transactions will be less than optimal. Land trans-
actions generally increase efficiency in resource allocation, as agents with high
(potential) marginal productivity of land are induced to acquire land from
agents with low marginal productivity.
In order to reduce the inefficiencies arising from uncertainty, societies develop
sophisticated institutional arrangements for recording and enforcing land rights.
One such arrangement is a centralized public record of land tracts and the
possessors of rights over these tracts. Such records have coverage at various
levels of geographic units (for example, county, provincial, or national), pre-
sumably with higher costs as the unit of coverage expands. As early as 600 B.C.,
the Bible (Jeremiah, chap. 32) describes that for a land transaction between the
prophet Jeremiah and a relative, two copies of the record of the transaction were
kept with a certain priest in the capital, Jerusalem. This arrangement gave
 Feder and Feeny 141

individuals who were considering buying or renting land from others a way to
verify that the rights they were about to purchase did indeed belong to the seller
or lessor. In later times, officially maintained land records and title documents
became much more systematic.
A central record is of course only one of the institutions designed to reduce
uncertainty. A functioning legal system and effective enforcement mechanisms
are necessary as well. In the absence of such public services, each individual will
increase his private allocation of resources for enforcement through the use of,
for example, guards or elaborate fences. It is more efficient to reduce the risk (at
least partially) through a public good (police, judiciary), than through individual
actions only.
Risk and asymmetric information with respect to land rights are particularly
extreme in frontier areas where land which was previously ownerless (it may
have been formally state-owned) is being claimed by individuals migrating from
other areas. In such circumstances there is no established community from
which knowledge can be obtained. The large number of claims and challenges to
claims typically overloads the administrative infrastructure (land record offices,
courts, and police), and it is not uncommon in such areas to find private (and
necessarily segmented) institutions to protect property rights over land (gunmen,
fortified properties).
Land Rights and Credit Transactions
The business of lending is inherently risky. The use of collateral on loans
reduces uncertainty and moral hazard problems for creditors. Collateral is more
valuable the more immobile and immune to damage it is, and land has tradi-
tionally been an ideal collateral asset in areas where land is scarce (Binswanger
and Rosenzweig 1986). The emergence of profit-motivated credit activities
(whether formal or informal) among agents within and outside established com-
munities is frequently an important element in inducing institutional change
with respect to land rights.
Land's usefulness as collateral is dependent on the absence of uncertainty and
asymmetric information with regard to the rights (in particular, transfer rights)
of the operator-occupier. A lender, for the same reasons which concern a poten-
tial buyer or renter, would like to be assured that the borrower-operator has
indeed the right to dispose of the land by sale or transfer or the right to transfer
use rights (a well-defined set of use rights over a sufficiently long time period has
a capitalized value which can serve as collateral). The availability of land as
collateral, and documentation of land rights which make such collateral cred-
ible, affect the willingness of creditors to make loans (Feder, Onchan, and
Raparla 1988). In addition, formal procedures for registering liens on property
rights provide important enforcement mechanisms. Thus the same institutional
arrangements that increase incentives for productive use of land also facilitate a
more efficient credit market.
Formal procedures, however, may also entail high transaction costs. There is
142 THE WORLD BANK ECONOMIC REVIEW, VOL. S, NO. 1

evidence of analogous informal mechanisms with lower costs. For instance, one
device is for the debtor to leave the physical title (document of land rights) on
deposit with the creditor. Although this does not provide for the formal registra-
tion of the transaction and therefore does not provide for a secure mechanism
for foreclosure in the event of default, it does give the creditor the ability to
ensure that the property is not disposed of without his interests being protected.
It also gives the creditor the ability to limit indirectly the total liabilities of the
debtor in that other formal (or informal) procedures requiring the presentation
of the title cannot be performed without the knowledge of the creditor. Formal
and informal practices of this nature have been observed in various countries in
Africa, and in India and Thailand (Meek 1946, p. 256; Stifel 1976). A contem-
porary informal variation which has emerged in Thailand recently is a signed
power of attorney agreement for the debtor to retain use of the land while
leaving the land title document with the creditor (Siamwalla and others 1990, p.
280). The availability of the land title document provides the creditor with
added security, making it possible for creditors to extend credit to persons with
whom they would otherwise not be sufficiently familiar. For the debtor these
informal arrangements provide access to larger sums at lower rates and reduce
market segmentation.
Public Sector Resources to Promote Land Security
In the rural areas of many developing countries the institutional arrangements
necessary to provide incentives and reduce uncertainty and asymmetric informa-
tion are often not well developed or are largely absent. This is not because the
forces which tend to generate these institutional arrangements are absent, be-
cause these forces are clearly present in many instances: high population-to-land
ratio, technology requiring fertility enhancement, and active or potential credit
markets in which property rights in land could serve as collateral. Rather, the
deficiencies stem from the overall inadequacy of public resources. The adminis-
tration dealing with land records may suffer from deficient technology (for
example, handwritten record retrieval methods when microcomputers would be
much more efficient), insufficient labor resources, and inappropriate storage
facilities. The judicial and police systems may be understaffed or underpaid,
which creates conditions for rent-seeking and for a slow process of property
rights enforcement.
In some countries the legal apparatus defining property rights may be exces-
sively complex and require various types of documents and affidavits which may
be useful in an urban context but not in an agricultural context. The complex-
ities increase the fixed transaction cost associated with enhancing the security of
property rights (for example, requiring the assistance of expensive lawyers and
demanding substantial time inputs from farmers.) This may create a stratifica-
tion whereby wealthier and larger farmers find it easier and more worthwhile to
finance these transaction costs (which tend to be relatively size-invariant),
whereas smaller and poorer farmers would not undertake them.
 Feder and Feeny 143

The inadequacy of public sector resources for reducing uncertainty in prop-

erty rights is aggravated by a public institutional framework that makes it diffi-
cult or impossible for private sector agents to substitute, if only in part, for the
lack of public infrastructure. In some countries, private land surveyors are not
recognized and certified, or the facilities to train such surveyors are not estab-
lished. The verification of boundaries and the resultant improvement in owner-
ship security is thus totally dependent on public sector land surveying, which is a
function of public budgets.
Such circumstances do not allow a significant reduction of the uncertainty
regarding property rights which the formal public sector institutions are sup-
posed to bring about. Instead, one frequently finds an emergence of localized
and informal risk-reducing institutional arrangements which may be less effec-
tive than a well-functioning formal institution but are better than no institution.
The segmented nature of these local arrangements implies a smaller volume of
transactions than that which would take place otherwise, and thus a less effi-
cient resource allocation.
A typical example is to be found in the western expansion period of the United
States (Anderson and Hill 1975; Dennen 1976). Under the provisions of home-
steading, ranchers in the Great Plains were able to claim as private property
parcels of land which were too small to support large herds of livestock.
Ranchers therefore supplemented grazing on private lands with grazing on the
public domain, which created the potential for conflicts.
Voluntary collective action in the form of cattlemen's associations helped to
ameliorate the situation. The cattlemen's association provided informal prop-
erty rights to the range, organized the spring round-up (an activity for which
there were economies of scale), and helped in the recognition and enforcement
of brands, a means for recording private property rights in livestock. These
mechanisms were especially important in the period before the availability of
barbed wire, a cheap means of enforcing rights to the range in a natural environ-
ment in which wood, stone, and other traditional fencing materials were scarce.

III. A MODEL OF INVESTMENT, PRODUCTION, AND LAND PRICE

DETERMINATION, AND SOME EMPIRICAL EVIDENCE

In this section, we summarize the results of a formal model which is presented

in the appendix. The model makes a number of simplifying assumptions, but it
nonetheless captures important characteristics of a rural economy in which land
rights are subject to risk. The link between land rights and the credit market is
incorporated through the assumption that the supply of credit available to
farmers is positively related to the value of their landholdings (as it would
typically be when land is a collateral for loans) and negatively related to their
probability of land loss.
The model assumes that the objective of farmers is to maximize their utility of
144 THE WORLD BANK ECONOMIC REVIEW, VOL. 5, NO. I

current consumption and next period's wealth by allocating their initial endow-
ment and borrowed funds to three uses: current consumption, land acquisition,
and investment in physical capital. Land and capital are used to produce their
next period's output through a neoclassical production function. Output and
land value, minus debt repayment (principal plus interest), make up the next
period's wealth. The risk to property rights is represented by a nonzero proba-
bility that land (and the output derived from the land) will be lost in the next
period to the present decisionmaker. The farmer thus perceives an expected
value of the next period's wealth that depends on the probability of land loss.
The model confirms that optimal allocation implies a negative relation be-
tween land price and the demand for land, as one would indeed expect in an
ordinary demand function. The capital-to-land ratio, however, is positively re-
lated to the price of land, because as land becomes more expensive, capital is
substituted for land (current consumption would increase as well). An increase
in the risk of land loss would, if land price were to be held constant, reduce both
the demand for land and the total demand for capital, as the output to be
produced by these factors has a lower probability of materializing, and because
the likelihood that land will remain part of the farmer's wealth declines. Another
reason for the decline in demand for these factors is that the supply of credit is
negatively affected by the increased risk to the viability of land as collateral.
Land prices, however, cannot remain unaffected by increases in the risk of
land loss, because as farmers reduce their demand for the fixed supply of land,
an excess supply is created which drives down the equilibrium price. Land prices
are thus shown to be negatively related to the riskiness of land rights. The
reduction in land values with increased risk diminishes further the supply of
credit per unit of land. Whereas the total amount of land employed in equilib-
rium is fixed, the reduced supply of credit per unit of land reduces the total
amount of capital acquired. As a result, at equilibrium the capital-to-land ratio
declines, and hence output per unit of land falls.
The model indicates that the equilibrium price of land contains a "collateral
premium" which is a result of the owner's ability to obtain additional and
cheaper credit by pledging the land as collateral to overcome the information
asymmetry in the credit market. This has an important implication for the
financing of land acquisitions: as the sales value includes the collateral premium,
the purchaser will not be able to pay for the land out of the benefit stream unless
he acquires it out of equity, at least in part. In the context of land reform,
landless beneficiaries with no equity cannot therefore be expected to compensate
former owners at full market price from the revenues of the farm.
Social welfare in one period is defined as the expected value of output minus
the value of real resources (capital) consumed in the process of production. The
analysis indicates that under ideal conditions (that is, no risk of land loss and an
interest rate which equals the rate of time preference) the price of land equals the
discounted value of the stream of social welfare generated by a unit of land.
With a nonzero risk to land rights, the price of land will be lower than its social
 Feder and Feeny 145

value (that is, the stream of net benefits generated by it). The reason for this
distortion is that the risk of losing land, aside from causing temporary output
reduction (which is a loss to both the individual and society), is also a risk of
asset loss to individuals. Society does not, however, risk a loss of the land, which
will remain a productive asset regardless of whether it is possessed by any
specific individual. This deviation will cause a difference between private and
social assessments of the benefits of eliminating the risk to property rights:
individuals will be willing to support a larger expenditure toward eliminating
uncertainty than is socially optimal.
A caveat is in order with respect to this result. The model assumes that
farmers are identical in their farming skills and differ only in their endowments.
In reality, however, there are differences among farmers in their farming skills.
This implies that the elimination of uncertainty, to the extent it expands the land
market, will bring social benefits by facilitating sales of land by individuals with
low productivity to ones with high productivity. This benefit may not be fully
reflected in private valuations of reductions in uncertainty and can thus intro-
duce a countervailing effect to the model's results.
Empirical results from a case study in Thailand confirm many of the proposi-
tions propounded in this and the preceding section. The study, reported in Feder
and others (1988a, 1988b), compared the performance of squatters on state
land, who lack titles on land they farm, with that of titled farmers. The results
show that titled land-rights to which had relatively little asymmetry in
information-bore little risk of expropriation, provided better access to credit,
and had a significantly higher market value as compared with squatters' land.
Titled farmers had a larger volume of investment, higher likelihood of land
improvements, more intensive use of variable inputs, and higher output per unit
of land.
A surprising finding was that most of the impact of title ownership in this
particular case stemmed from the fact that titles increased farmers' access to
formal credit, rather than from elimination of actual risk to the land rights of the
farmers. For example, in a comparison of two groups of squatters, one of which
was granted official acceptance but not permission to legally transact in land, no
difference in performance was apparent. Both groups were, of course, unable to
pledge their land as collateral for a debt. Similarly, of three areas studied, in the
area in which the dominant source of credit was the informal credit market (in
which information asymmetry is small as compared with the formal credit mar-
ket), the differences between land values and agricultural performance of titled
and untitled farmers were the smallest. As confirmed through other sources, the
risk to squatters' land rights in Thailand is not significant because eviction by the
state would be politically costly. Formal credit institutions, however, do not
accept land as collateral without formal title for reasons related to information
asymmetry. A calculation over all survey areas of the costs and benefits of
providing squatters with legal title ownership shows that the benefits outweigh
the costs by a wide margin.
 146 THE WORLD BANK ECONOMIC REVIEW, VOL. 5, NO. I

IV. CONCLUSIONS

The definition and institutionalization of property rights in land have been an

important issue for societies throughout history. The nature of these rights and
the way they are enforced have significant consequences for resource allocation
and economic efficiency. Changes in population density, technology, and politi-
cal power generate changes in the assignment of property rights and in the
institutional arrangements associated with these rights. Generally, secure indi-
vidual property rights over land, or secure and long-term use rights on land
induce exertion of higher levels of labor and management effort and higher
levels of investment to protect or enhance land fertility. There are exceptions to
this proposition, such as cases in which unregulated individual use generates
externalities. The provision of secure individual property rights requires not
only social rules for allocation of land rights, such as a constitution which
recognizes individual land rights, but also adequate implementation and en-
forcement mechanisms. In the absence of such mechanisms, uncertainty regard-
ing land ownership rights will generate inefficiencies in the allocation of re-
sources. The inefficiencies are aggravated if the use of land as collateral affects
credit availability. In this case, the same institutions which improve the effi-
ciency of the land market will also increase efficiency in credit markets.
Asymmetry in information regarding the allocation of land rights between
agents transacting in land is another potential source of inefficiency which is
likely to increase with agricultural development. The asymmetry promotes the
segmentation of land markets, constraining the volume of transactions and thus
hindering efficient resource allocation. Various institutional arrangements were
developed over time to reduce the extent of asymmetric information, such as
land records and title documents. These arrangements typically require public
sector involvement because of economies of scale and complementarity with
other public sector activities-for example, cadasters for taxation purposes and
judicial and police functions.
Although there are obvious social benefits to the provision of secure property
rights and to the removal of information asymmetry, there is a likelihood that
private benefits of secure land ownership rights exceed the contribution of such
security to society's resources. Such a situation may imply that individuals are
willing to support policy measures to improve ownership security to a point at
which marginal costs exceed the marginal gain in net output. Appropriate taxes
or user charges can be used when implementing some of the institutional ar-
rangements to enhance the security of land rights, so as to bring individuals'
willingness to support such programs in line with the net gains accruing to
society as a whole. The gains in land and credit markets' efficiency which result
from reduction of uncertainty in land rights, however, may generate externalities
countervailing the excess valuation of security by individuals and, in any case,
are a social benefit of considerable magnitude in economies where these markets
are relatively free. The total social benefits of improving security of land rights
 Feder and Feeny 147

in such economies are therefore likely to exceed social costs, although the exact
optimal extent of security enhancement (and the associated costs) will require an
assessment based on the specific situation being considered. But in areas where
credit and land markets are not yet developed, an investment in titling and land
registration may entail an excessive cost in comparison with the benefits, and
security of tenure can be enhanced by cheaper methods such as legalizing the
authority of local institutions.

APPENDIX: A MODEL OF INVESTMENT, PRODUCTION, AND LAND PRICE

DETERMINATION

For simplicity assume a two-period horizon, in the first period of which land
acquisition, consumption, and investment decisions are made that determine
production in the second period. Capital is completely used up in the process of
production. Capital is the numeraire good, with price 1, and is available with
infinite supply elasticity to the rural sector. Individuals maximize an expected
utility function which is separable into two arguments: current consumption and
terminal wealth. A further simplification is that the utility function is linear in
terminal wealth. Risk to property rights is introduced through a nonzero proba-
bility X that the land and its second-period output will be taken from the current
decisionmaker, for example, through takeover by other individuals by force or
legal challenges. The possibility of gaining land through such actions is viewed
as an exogenous probabilistic event. Although the benefits of such a windfall
should enter the objective function, it can be shown that this element does not
affect the results of the model, and for simplicity it is not included explicitly.
The notation used in the model is: T = land, P = price of land, k capital-
to-land ratio, CO = first-period consumption, WO= initial wealth, k = proba-
bility of ownership loss, U = utility of first-period consumption, and y = output
per unit of land.
Model Components
The production function exhibits constant returns to scale in land and capital.
The per hectare output is therefore
(A-1) y = y(k); y' > 0; y" < 0
Utility of current consumption is a concave function with decreasing marginal
utility:
(A-2) U= U(Co); U' > 0; U" < 0
Individuals maximize their expected utility, which is composed of the utility of
current consumption plus their expected terminal wealth. Terminal wealth is
equal to output plus land value in the case that land rights are not lost. Maximiz-
ation is subject to a budget constraint whereby the value of land, capital, and
current consumption cannot exceed initial wealth plus borrowed funds.
148 THE WORLD BANK ECONOMIC REVIEW, VOL. 5, NO. I

It is assumed that credit is rationed, and that the ration is binding for all
farmers. The ration is proportionate to a borrower's landholding value with
land serving as collateral. Denote the proportion by s. The total credit ration
(say, S) is positively related to land's ownership security, that is:
(A-3) S = s(O)PT: s' < 0; 0< s< 1
A fixed rate of interest (r) is assumed. The model could be developed with an
assumption of an interest rate dependent on risk to property rights, which would
yield even stronger results. Because all farmers are assumed to be rationed, the
marginal productivity of capital is necessarily higher than the cost of credit.
If land rights are lost in the second period, the farmer is still obliged to pay the
debt acquired in the first period. (It could be assumed that only a proportion of
land is lost, say, -y, and that farmers repay their debts from their remaining
wealth. In this case, the results of the model would be practically unchanged,
with the term -yo replacing 4 in all derivations. For simplicity, the calculations in
the text assume -y = 1. Even though all debt is ultimately repaid, lenders are
concerned about the risk of land loss because of the transaction cost of collecting
debts from dispossessed farmers. This motivates the assumption that the credit
ration depends on land security, as in (A-3).
The farmer's objective function is
(A-4) max U(CO)+ (1 - O)Tfy(k) + P] - (1 + r)s(O)PT
CI),T k

subject to the budget constraint

(A-5) WO+ s(O)PT = kT + PT + Co
From (A-5) one obtains
(A-6) CO = WO-kT-PT + s(O)PT
Thus, the objective function can be written
(A-7) max U[Wo - PT(1 - s) - kT] + (1 - O)T[y(k) + P] - (1 + r)s(4)PT
T,k

First-order conditions for a maximum are

(A-8) (1 - )(y + P) -U' [P(l1-s) + k - (1 + r)s(X5)P]
=O

(A-9) (1 T - TU' = 0
The Hessian matrix is
s) + k] 2
AU"[P(1 - U"[P(1 - s) + k]T
(A-10) [HI = LU"[P(1- s) + k]T T(1 - 4)y" + T2U" I
Second-order conditions are clearly satisfied as the determinant is positive:
(A-11) IHI -- A = T(1 - k)U"[P(1 - s)P + k] 2 y" > 0.
(because U" < 0, y" < 0).
 Feder and Feeny 149

[_ I
The impact of a change in land price is given by:

[d O -(1
k) - - U"[(l - s)P + k](1 -s)T
(A-12) [H] dk -T 2 U"(i-s)

where use has been made of equations (A-8) and (A-9). The concavity of y
implies y > y' k.
Using Cramer's rule, one obtains:

(A-13) dT= 1- (1 (y Y' ) [T(I - O)y" + T 2 U"]

- U"[(1 -s)P + k]T 2 (1- s)(1 - O)y" | < 0

That is, the demand for land is negatively related to its price, as intuitively
expected. The sign can be established by noting the concavity of y and U. From
Cramer's rule, one also obtains

(A-14) dk = (1- ) (y - y ) U"[(1 -s)P + k] T > 0

d_P P A
That is, the capital-to-land ratio increases with land price. This is also intuitively
expected, because with higher land prices, farmers will substitute capital for
land in production.
Because the demand for land is monotonically and negatively related to the
price of land, and given that the supply of land is fixed, there is a stable equilib-
rium price of land which depends on the parameter ¢>,the risk to ownership.
Intuition suggests that an increase in the probability of loss of land rights should
reduce the demand for the risky asset (land) if land prices are fixed. This is
confirmed by the following derivation:

dT Y + P-[(1-4)y' -(1 + r)]Ps' +

(A-15) [H] dk = TU" [(1 - s)P + k]Ps'
LdkLTy' + T 2 U"Ps'
Denote the first element in the vector on the right-hand side of (A-15) by 0,
and note 0 > 0, because (1 - 0)y' - (1 + r) > 0 if the credit constraint is
binding. Then, using Cramer's rule and equations (A-8) and (A-9)

(A-16) dT = 1 {OT(1- )y" + [T 2 (l + r)sPU +

-T 2 U"[(l + O)y' - (1 + r)]Ps' I < 0
dk 1 -Oy )
(A-17) To = A I[(1- )y' -(1 + r)]
Ps' - [(1 + r)sP/(l - 0)]}TU" > 0
Thus an increase in the risk of land loss would increase the capital-to-land
150 THE WORLD BANK ECONOMIC REVIEW, VOL. 5, NO. I

ratio if land price were held constant. This result obtains even though the
relative prices of credit and land are held constant, because the incentive to buy
land due to its credit-enhancing role (aside from its productive contribution) has
been diminished, and therefore the ratio of marginal contributions of land and
capital has changed. The overall demand for capital, however, will decline with
an increase in risk to land rights even when land prices are fixed, because the
higher risk reduces the expected return on investment. This can be verified by
calculating
d(kT) _ Tdk kdT
do - d d

utilizing equations (A-16) and (A-17).

While the capital-to-land ratio increases with risk when land price is fixed, the
relation is reversed in equilibrium. An increase in ownership risk reduces the
demand for land (dT/do) < 0), whereas the supply of land is fixed, so that the
price will decline to equate demand to the fixed supply. Using equations (A-13)
and (A-16), one can show that the price of land in equilibrium will decline,
(dP/dO < 0).
(A-18) dP =- {OT(1 - O)y" + [T 2 (1 + r)sPU"/(1 + d)]
do - T 2 U"[(1 - O)y' - (1 + r)]Ps' }{(1 - 0)[(y - ky' )/P]
T(l - O)y" + T 2 U"
- U"[(1 - s)P + k]T 2 (1 - s)(1 - O)y"1l < 0

In this model, in which credit availability depends on the riskiness of the land
collateral, the impact of higher ownership risk on the price of land has two
components. The first term of the right-hand side of (A-18) reflects the impact of
risk on the farmer's resource allocation: as uncertainty increases, present con-
sumption is preferred to future wealth accumulation, and the demand for land is
reduced while the supply is fixed, which requires a price reduction to restore
equilibrium. The second and third terms on the right-hand side of equation
(A-18) reflect respectively the impact of land ownership and ownership risk on
the supply of credit: as uncertainty rises, the farmer's access to credit for land
purchases or capital diminishes and the demand for land is reduced. If there
were no credit market (s = 0), or if credit supply was not affected by ownership
risk (s' = 0), then these components would vanish, but land value would still be
negatively affected by higher ownership risk. These results also demonstrate the
link between land price and land's role as a collateral. The price of land includes
a premium reflecting the additional income due to the credit which can be
acquired by pledging the land, and which in turn increases, at the margin, the
farmer's utility.
The change in the equilibrium capital-to-land ratio following an increase in
risk and the subsequent reduction in land price can be calculated (using equa-
tions (A-12), (A-14), (A-16), (A-17), and (A-18)) as:
dk -* dk dk dP _ dk dk [dT/d+]
dq5 d= dP d = dk dP [dT/dP]
 Feder and Feeny 151

(A-19) = ({[Ty' (1 - 0) (y y' k)/P] + TU"(r + i)s'

- + T 2 U" (1 - s)
{[(1 - s)P + k]y' -(y + P)})
[(dT/dP)A]-l < 0

where denotes equilibrium value after market adjustments have taken place.
Equation (A-19) confirms that equilibrium capital-to-land ratios decline as a
result of higher uncertainty. The intuition is that higher ownership uncertainty
increases current consumption at the expense of demand both for land and
capital goods. But the price of land declines to clear the market at the original
level of land use as the supply of land is fixed. All of the decline in the purchase
of investible resources is thus absorbed by the capital good, reducing capital-to-
land ratios. The increase in uncertainty thus causes a decline in output per unit
of land.
Valuation of Policies to Eliminate Uncertainty in Land Rights
Suppose that whenever property rights are challenged, the production process
is interrupted and one period's output is lost. Then the expected net benefit to
society from a unit of land in one period is [1 - k]y - U' k, where the capital
stock consumed in the process of production is evaluated in terms of its marginal
welfare opportunity cost (U' ). Denote U' = 1 + 6, where 6 may be viewed as a
time preference premium, because one unit of second-period wealth yields U'
units of utility if transferred to the first period. Then, rearranging equation
(A-8), one obtains
(A-20) (1 -)y - (1 + 6)k = [(r-6)s +4 + 6]P
Define net social benefits as the value of net addition to economic resources
(that is, output less the real value of resources used in production). Integrating
both sides of equation (A-20) over an infinite time horizon and using 6 as a
discount factor, the total discounted net social benefits generated by a unit of
land are

(A-21) |[(1 - k)y - (1 + 5)k]e-atdt 1+ _ (6- r)SI

r p

When there is no risk and no credit market distortions (such that 4 = 0, 6 =

r), the right-hand side reduces to P. The price of land reflects exactly the net
social benefits it generates. In the presence of risk and distortions, the price of
land differs from the net social benefits. One reason is that the possibility of
asset ownership loss is a risk of capital loss to individuals, but not to society,
which risks only one period's output. The second source of deviation comes
from credit rationing. When credit is priced below the opportunity cost of
capital (U' > 1 + r), the value of the subsidy is capitalized in the land value.
To evaluate private and social benefits from policies and institutions eliminat-
ing uncertainty in property rights, denote P,, as the price prevailing when the
probability of land loss is X, while P0 is the price when there is no risk. Also,
assume for simplicity that there are no credit market distortions (r = 6). The
benefit to an individual (per unit of land) from institutional arrangements that
152 THE WORLD BANK ECONOMIC REVIEW, VOL. S, NO. I

eliminate the risk of losing property rights can be expressed as the difference
between land prices without and with risks (Po -PO).
Denoting the left-hand side of equation (A-21) as B.,, the benefits to society
(the discounted expected addition to resources) from elimination of risk to
property rights in land are given by

(A-22) Bo - B6 = (PO - P6) - o

The net addition to society's resources is thus smaller than the benefits to
individuals.

REFERENCES

Alchian, Armen A., and Harold Demsetz. 1973. "The Property Rights Paradigm." Jour-
nal of Economic History 33 (March): 16-27.
Anderson, Terry L., and P.J. Hill. 1975. "The Evolution of Property Rights: A Study of
the American West." Journal of Law and Economics 18 (April): 163-79.
Arrow, Kenneth. 1985. "Maine and Texas." American Economic Review 75 (May):
320-23.
Berkes, Fikret, David Feeny, Bonnie J. McCay, and James M. Acheson. 1989. "The
Benefits of the Commons." Nature 340 (July): 91-93.
Binswanger, Hans, and Mark Rosenzweig. 1986. "Behavioral and Material Determi-
nants of Production Relations in Agriculture . Journal of Development Studies 22
(April): 523-39.
Campbell, Bruce, and Ricardo A. Godoy. 1986. "Commonfield Agriculture: The Andes
and Medieval England Compared." In National Research Council, Proceedings of the
Conference on Common Property Resource Management. Washington, D.C.: Na-
tional Academy Press.
Coase, Ronald H. 1960. "The Problem of Social Cost." Journal of Law and Economics 3
(October): 1-44.
Dennen, R. Taylor. 1976. "Cattlemen's Associations and Property Rights in Land in the
American West." Explorations in Economic History 13 (October): 423-36.
Engerman, Stanley. 1973. "Some Considerations Relating to Property Rights in Man."
Journal of Economic History 33 (March): 43-65.
Feder, Gershon, and Raymond Noronha. 1987. "Land Rights Systems and Agricultural
Development in Sub-Saharan Africa." World Bank Research Observer 2 (July): 143-
69.
Feder, Gershon, Tongroj Onchan, Yongyuth Chalamwong, and Chira Hongladarom.
1988a. Land Policies and Farm Productivity in Thailand. Baltimore, Md.: Johns
Hopkins University Press.
Feder, Gershon, Tongroj Onchan, and Yongyuth Chalamwong. 1988b. "Land Policies
and Farm Performance in Thailand's Forest Reserve Areas." Economic Development
and Cultural Change 36 (April): 483-502.
Feder, Gershon, Tongroj Onchan, and Tejaswi Raparla. 1988c. "Collateral, Guaranties
and Rural Credit in Developing Countries: Evidence from Asia." Agricultural Eco-
nomics 2 (November): 231-45.
 Federand Feeny 153

Feeny, David. 1982. The Political Economy of Productivity: Thai Agricultural Develop-
ment 1880-1975. Vancouver, B.C.: University of British Columbia Press.
1988a. "The Demand for and Supply of Institutional Arrangements." In Vincent
Ostrom, David Feeny, and Hartmut Picht, eds., Rethinking Institutional Analysis and
Development: Issues, Choices, and Alternatives. San Francisco, Calif.: Institute for
Contemporary Studies Press.
1988b. "The Development of Property Rights in Land: A Comparative Study."
In Robert H. Bates, ed., Toward a Political Economy of Development: A Rational
Choice Perspective. Berkeley: University of California Press.
Hallowell, A. Irving. 1943. "The Nature and Function of Property as a Social Institu-
tion." Journal of Legal and Political Sociology 1 (April): 115-38.
Libecap, Gary D. 1986. "Property Rights in Economic History: Implications for Re-
search." Explorations in Economic History 23 (July): 227-52.
Meek, C. K. 1946. Land Law and Custom in the Colonies. London: Oxford University
Press.
North, Douglass. 1981. Structure and Change in Economic History. New York: Norton.
Roumasset, James A., and Sumner J. LaCroix. 1988. "The Coevolution of Property
Rights and Political Orders: An Illustration from Nineteenth-Century Hawaii." In
Vincent Ostrom, David Feeny, and Hartmut Picht, eds., Rethinking Institutional
Analysis and Development: Some Issues, Alternatives, and Choices. San Francisco,
Calif.: Institute for Contemporary Studies Press.
Schlatter, Richard. 1951. Private Property: The History of an Idea. New Brunswick,
N.J.: Rutgers University Press.
Siamwalla, Ammar, Chirmsak Pinthong, Nipon Poapongsakorn, Ploenpit Satsanguan,
Prayong Nettayarak, Wanrak Mingmaneenakin, and Yuavares Tubpun. 1990. "The
Thai Rural Credit System: Public Subsidies, Private Information, and Segmented Mar-
kets." World Bank Economic Review 4, no. 3 (September): 271-95.
Solow, Robert M. 1985. "Economic History and Economics." American Economic
Review 75 (May): 328-31.
Stifel, Laurence D. 1976. "Patterns of Land Ownership in Central Thailand during the
Twentieth Century."Journal of the Siam Society 64 (January): 237-74.
Taylor,John. 1988. "The Ethical Foundations of the Market." In Vincent Ostrom, David
Feeny, and Hartmut Picht, eds., Rethinking Institutional Analysis and Development:
Issues, Alternatives, and Choices. San Francisco, Calif.: Institute for Contemporary
Studies Press.
Umbeck, John F. A. 1977. "The California Gold Rush: A Study of Emerging Property
Rights." Explorations in Economic History 14 (July): 197-226.
Wade, Robert. 1986. "Common Property Resource Management in South Indian Vil-
lages." In National Research Council, Proceedings of the Conference on Common
Property Resource Management. Washington, D.C.: National Academy Press.
View publication stats