Journal of Development Economics 96 (2011) 66–78

Contents lists available at ScienceDirect

Journal of Development Economics

j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / d eve c

Land tenure differences and investment in land improvement measures: Theoretical

and empirical analyses
Awudu Abdulai a,⁎, Victor Owusu b, Renan Goetz c
a
University of Kiel, Kiel, Germany
b
Kwame Nkrumah University of Science and Technology, Kumasi, Ghana
c
University of Girona, Girona, Spain

a r t i c l e i n f o a b s t r a c t

Article history: This article develops a theoretical framework to examine the relationship between land tenure arrangements
Received 16 July 2008 and households' investment in soil-improving and conservation measures. It then analyzes this relationship
Received in revised form 26 July 2010 with a multivariate probit model based on detailed plot-level data from villages in the Brong Ahafo region of
Accepted 1 August 2010
Ghana. A major hypothesis tested is that investment in productivity-enhancing and conservation techniques
are influenced by land tenure arrangements. The theoretical analysis and empirical results generally reveal
JEL classification:
O13
that land tenure differences significantly influence farmers' decisions to invest in land-improving and
O55 conservation measures. The findings also show that tenure security does affect farm productivity.
Q12 © 2010 Elsevier B.V. All rights reserved.
O1

Keywords:
Land tenure
Property rights
Investment
Optimal control
Farm productivity

1. Introduction A central issue of the related empirical investigations is the effect of

tenure security on investment and productivity. On theoretical
The role of land tenure on investment in productivity-enhancing grounds, three main arguments have been advanced for a positive
measures in developing countries has been widely documented in the link between tenure security and investment. First, secured property
economic literature. Since land is central to the social and economic rights are expected to provide a guarantee for farmers to undertake
development of a vast majority of the people living in Sub-Saharan long-term investment in land-improving and conservation measures,
Africa, the link between indigenous tenure arrangements and since there would be no fear of expropriation. As noted by Banerjee
productivity-enhancing investments has attracted the attention of and Ghatak (2004), given that the result of land-improving invest-
both researchers and policy makers. While studies by Dorner (1972) ments is normally realized with a one period lag, if the tenant is evicted
and Harrison (1987) argued that indigenous tenure systems provide with some possibility during this period, he will be enjoying only a
insufficient security to induce farmers to undertake soil-improving fraction of the benefit from the investment in expected terms. This
investments, Noronha (1985) pointed out that these arrangements may cause the tenant to supply a lower level of investment in effort for
are dynamic and evolve in line with factor prices. The significance of the same crop share, a reason why security of tenure is thought to be
this debate has attracted a great deal of attention among economists good for investment. Second, it has been argued that secured land
(Binswanger and Rosenzweig 1986; Besley, 1995; Quisumbing et al., rights make it easier to use land as collateral to obtain loans to finance
2001a; Brasselle et al., 2002; Place and Otsuka, 2002; Jacoby et al., agricultural investments (Feder and Feeny, 1991). The third effect
2002; Banerjee and Ghatak, 2004; Goldstein and Udry, 2008; operates through better possibilities for trade. If improved transfer
Deininger and Ali, 2008). rights enhance factor mobility by making it easier for farmers to sell or
rent their land, investment in land-improving measures may be
facilitated. An issue that has gained increasing significance in recent
⁎ Corresponding author. Department of Food Economics and Consumption Studies,
University of Kiel, Olshausenstrasse 40, 24118 Kiel, Germany. Tel.: + 49 431 880 4426;
empirical analysis is the endogeneity of land rights in estimating the
fax: + 49 431 880 7308. effect of tenure security on agricultural investment. Authors like
E-mail address: aabdula@food-econ.uni-kiel.de (A. Abdulai). Besley (1995), Quisumbing et al. (2001a), Place and Otsuka (2002),

0304-3878/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.jdeveco.2010.08.002
 A. Abdulai et al. / Journal of Development Economics 96 (2011) 66–78 67

Brasselle et al. (2002) have rightly noted that farmers may undertake results are discussed in Section 6, while the final section presents
land-improving investments in order to gain tenure security.1 some concluding remarks.
As rightly noted by Besley and Ghatak (2009), the empirical
findings from the land rights–investment relationship appear to be
inconclusive. Studies on Africa by Migot-Adholla et al. (1994) and 2. Land tenure in Ghana
Pinkney and Kimuyu (1994) reveal that the impact of land rights on
soil-improving investment and planting of tree crops is quite low. As in several other African countries, land is traditionally owned by
On the other hand, studies by Besley (1995) on Ghana, Jacoby et al. the community in Ghana. Control of the land is transmitted through
(2002) on China (2002), as well as Carter and Olinto (2003) on the elders, who are custodians of the land. Each headman therefore
Paraguay show that tenure security exerts a positive and significant sees to it that all members of his lineage have portions to farm (Gildea,
impact on investments. Banerjee et al. (2002), who found a positive 1964). With the development of the cocoa industry in the country,
impact of tenure reform on farm productivity in India, partly practices of landholding have become individual ownership in contrast
attributed this to higher investment due to improved tenure to family control of segments of the community land. Particularly in
security. Brasselle et al. (2002) report that land tenure security is the Ashanti and Brong Ahafo regions, the procurement of large bundles
influenced by investment, and that once the endogeneity bias is of land by wealthy investors changed the old order. These investors
properly controlled, increased land rights do not appear to stimulate either moved into previously unclaimed land or acquired secured
investment. Place and Otsuka (2002) also found that coffee planting rights to community land in exchange for money or influence. Some of
is used by farmers to enhance tenure security, supporting the notion these large-scale farmers sometimes reside somewhere else and
that farmers consider tenure implications when making investment supervise the operations on their land (Benneh, 1989).
decisions. Even among migrants, land rights have become more clearly
The more recent studies appear to be showing positive impacts of individualized, with members of the family qualifying for inheritance
tenure security on investment. Deininger and Ali (2008) show that of land in the event of the death of the family head (Quisumbing et al.,
full land ownership, compared to mere occupancy rights, exerts a 2001b). There is a complex system of communally owned land in the
statistically significant and economically large effect on investment rural northern regions of the country, with many local variations. Land
and productivity of land-use in Uganda. Goldstein and Udry (2008) tenure is generally based on the community's social organization, and
also find that insecure land tenure in Ghana is associated with greatly the basic unit of ownership is the family or clan. Institutions of land
reduced investment in land fertility, while Jacoby and Mansuri (2008) tenure are now evolving towards individualized ownership through
find in their study on Pakistan that farmers invest less in their leased investments in tree planting and management, transfer of land as
plots than they do in their owned plots. grants and gifts, and the coming into effect of the Interstate Succession
This article contributes to the tenure-security-investment debate Law (PNDC 111) in 1985 (Otsuka et al., 2003). The law is a legal
by developing a framework that captures the impact of different land framework that provides equal rights of inheritance between spouses
tenure arrangements on investment decisions of farmers. The model and increased rights for children.2
embodies behavioral assumptions consistent with investment deci- Given that full ownership of rights over land traditionally resides
sions that characterize investment in productivity-enhancing inputs with the community, one becomes less concerned with overall land
in the agricultural sectors of most sub-Saharan African countries. First, tenure security than with rights that the individual holds over specific
we develop a model to examine the effects of 4 different land tenure land parcels (Place and Hazell, 1993). We therefore focus on the long-
arrangements on investment decisions of farmers on theoretical term interests farmers have on land parcels, in terms of their rights to
grounds. The investments include planting trees, mulching, and cultivate the land for long periods of time and their ability to rent or
application of organic manure and mineral fertilizers. We then use sell the land. As argued by Place and Hazell (1993), these features of
variations in tenure arrangements between 560 plots obtained from a land control are best captured by tenure measures based on the
survey of 246 farmers from 6 villages in the Brong Ahafo region of individual use and transfer rights that farmers possess over land. To
Ghana to analyze the impact of land tenure arrangements on capture these features, we collected detailed information on individ-
investment in soil-improving and conservation measures. The ual rights – basically use rights and transfer rights – for each parcel
empirical part of the article also examines the relationship between operated by the farmers in the sample.3
land tenure arrangements and farm productivity. The four main types of land tenure arrangements identified in
The main contributions of the article reside in the fact that the the survey area are owner-operated with full property rights,
results from the theoretical analysis hold for a wide range of owner-operated with restricted property rights, fixed-rent and
situations and are as such independent of case specific data. The sharecropping contracts. The owner-operated with full rights
empirical analysis considers a) endogeneity between land rights involves farmers owning and cultivating their own plots. Farmers
and investment decisions and b) interdependence between the cultivating these parcels have transfer rights, including rights to sell
different investment decisions. Our empirical evidence shows that the parcels, although in some cases family approval has to be
land tenure differences significantly influence farmers' decisions to acquired before the land can be sold. Owner-operated with
invest in land-improving and conservation measures, and that restricted rights involves plots that are acquired as grants, but
tenure differences do affect farm productivity, even after accounting cannot be transferred or inherited, although they may be rented out.
for household fixed effects. The fixed-rent arrangement involves land owners renting out
The rest of the paper is organized as follows. Section 2 discusses parcels to tenants, who are normally migrants from other areas.
land tenure in Ghana. In Section 3, we present a theoretical model on Under a sharecropping contract, an arrangement is made between
soil capital and forest use on plots where farmers can undertake short- the landlord and the operator, such that part of the output is given
term and long-term investments in land improvements. In Section 4, to the landlord as compensation for using the land. The two forms
the estimation and identification strategy employed in the empirical
analysis is outlined. Section 5 discusses the survey data. The empirical
2
Otsuka et al. (2003) have observed that instead of the stipulations of the law, some
local communities prefer a formula that gives one-third of the property to each spouse,
1
Quisumbing et al. (2001a) document in their study that when primary forests children and maternal family.
3
were abundant in the Akan area in Ghana, they were usually appropriated by young, As rightly noted by an anonymous reviewer, the more commonly used term for a
unmarried males for food-crop production. In return, they obtained relatively strong piece of land with a unique mode of land acquisition is parcel. Given that some authors
land rights for the substantial labor input that was required to clear forests. refer to it as plots, we use the terms synonymously in this paper.
68 A. Abdulai et al. / Journal of Development Economics 96 (2011) 66–78

are abunu (one-half of the output for the tenant and one-half for the the soil, and also protect agricultural crops against water and wind
owner) and abusa (one-third for the tenant).4 erosion. In addition, standing trees may also provide fruits. The fruit
yield is assumed to be proportional to the biomass of the trees, and is
3. Theoretical analysis specified as ϑ W(t). However, planting trees reduces the acreage that is
available for crop production. This reduction in acreage can be
The model that will be presented later analyzes the link between expressed as α W(t) with α N 0. The size of the entire farm is normalized
land tenure arrangements and investment in land-improving and to one and the share of the land that is used for crop production is
conservation measures within a dynamic framework. The previous denoted by L(t). Under the restriction that 0 ≤ L(t) = 1 − α W(t), W = 0
literature, considers standing forest (Angelsen, 1999; Barbier, 2004) implies that the entire land will be used for crop cultivation, whereas
or soil capital (Ehui et al., 1990) as a renewable or non-renewable W = 1/α implies using the entire land for growing trees.
resource. In this article, we model soil capital and trees as a Since current decisions tend to affect the evolution of the natural
renewable resource and analyze their interdependencies with resources over time, we analyze the farmer's decision problem within
agricultural production. It is assumed here that farmers combine a dynamic context and take into consideration the fact that the
investments in both mineral fertilizers, XM(t), such as NPK and planning horizon of the farmer depends on the land rights. We further
organic fertilizers, XOr(t), such as mulch and organic manure where assume that agricultural households maximize farm net benefits
t indicates calendar time.5 We control for cultivated plots and plots subject to agronomic and biophysical constraints (the evolution of the
whose areas of production have been used for tree planting. soil and forest) over a planning horizon of length tk − t0 where t0
Farmers are also assumed to choose production methods that denotes the initial point and tkthe final point of the planning horizon.
improve soil fertility and increase productivity. Although crop yields The residual value of the trees and soil capital is given by r(S(tk),
normally increase with higher rates of mineral fertilizer, yields may W(tk)). The function r(⋅)will be zero for fixed-rent tenants and
decline with time, if no organic fertilizer or any other organic sharecroppers, because they do not have the possibility to sell the
material is added. The decline in yields may result from soil land. Owner-operated with restricted property rights have the
degradation, which then erodes the original purpose of investment. possibility to rent out the land, but not sell or bequeth it, so that the
Given the potential physical and chemical degradation of the soil value of the function will be strictly positive. However, it is smaller
as a result of continuous application of mineral fertilizer, profit than that of the owners with full property rights because they can sell
maximizing farmers normally invest in organic fertilizers that build or bequeth the land. Given that restricted property rights are usually
up the soil structure and naturally replenish nutrients in the soil with not limited over time, it is assumed that owners with full or with
relatively low cost. Moreover, the nutrients supplied by organic restricted rights have the same long-term perspective,7 while tenants
fertilizer are available over a longer time horizon compared to (fixed-rent or sharecropping) have a planning horizon that corre-
nutrients supplied by mineral fertilizer (Besley, 1995; Jacoby et al., sponds to the stipulated tenure duration. Given these assumptions,
2002).6 Let us assume now that the production function is defined for the farmer's decision problem can be stated as
1 ha. Under this assumption, the agricultural production function per
tk
hectare can be defined as f(S(t), XM(t), XOr(t)), where S(t) represents −φt
soil capital, and XM(t) and XOr(t) are as defined previously. The max J ≡∫e ½ðpf ðSðt Þ; XM ðt Þ; XOr ðt ÞÞ−pM XM ðt Þ ð1Þ
L; XM ; XOr ; P;C
t0
application of organic fertilizers augments soil capital according to the
−pOr XOr ðt ÞÞLðt ÞÞ−pL ð⋅Þ + pW C ðt Þ + pF ϑ W ðt Þ
function h(XOr(t)), with h′(⋅) N 0. Moreover, since the mineral and
organic fertilizers are “perfect” substitutes in the short run, we can −φtk
−pP P ðt Þdt + e r ðSðtk Þ; W ðtk ÞÞ
write the function f(⋅)as a sum of expressions that reflect individually
the effect of XM(t), and XOr(t). Hence, we can assume that the cross
subject to
derivative, fXM XOr is equal to zero.
In contrast, soil capital and fertilizers are subsitutes to a lower
degree. As explained previously, the continuous application of XM(t), or Ṡðt Þ = hðXOr ðt ÞÞLðt Þ + yðW ðt ÞÞ−δf ðSðt Þ; XM ðt Þ; XOr ðt ÞÞLðt Þ;
XOr(t) has opposite effects on the evolution of S(t) over time. An
with Sðt0 Þ = St0 ; Ẇ ðt Þ = gðW ðt ÞÞ−C ðt Þ + P ðt Þ;
increase in S(t) results in a decrease in the marginal productivities of
XM(t) and XOr(t), while a decrease in S(t) yields the opposite effect.
with W ðt0 Þ = Wt0 ; 0≤Lðt Þ = 1−α W ðt Þ and XOr ðt Þ; XM ðt Þ; C ðt Þ≥0;
Therefore, we assume that fXMS b 0 and fXOr S b 0. The volume of the
biomass of trees (wood) is given by W(t). The farmers have the choice
to plant young trees and to cut trees, whose volume is denoted by P(t) where pL ð⋅Þ = ð1−θÞ p̄L + θγ1 pf ð⋅ÞLðt Þ + θγ2 pW C ðt Þ represents the
and C(t) respectively. The planted trees grow according to the logistic cost of the land cultivated with agricultural crops and planted trees,
growth function g(W(t)), with g′(⋅) N 0. Standing trees increase soil with θ = 0, 1. In the case of sharecropping θ = 1 and pL =γ1pf(⋅)L(t) +
capital specified by the function y(W(t)), with y′(⋅) N 0. This is because γ2pWC(t)), where γ1 and γ2 indicate the share of the yields that accrue
tree management practices can improve the availability of nutrients in to the owner of the land.8 In the case of no sharecropping (owner and
fixed-rent tenant), θ = 0, the cost of the land is given by the constant
p̄L , which denotes the annual land rent in the case of a tenant, or the
4
annual cost of capital (own and borrowed) in the case of an owner.
As noted by Jacoby and Mansuri (2008), landlords normally offer fixed-rent
contracts only to tenants with sufficiently high wealth. For tenants with binding
All parameters, except t0 , tk and St0 are grouped in a vector named ῡ.
wealth constraints, landlords normally go into share contracts. However, it is The components of this vector are given by p = price of the cultivated
significant to note that share-cropping contracts usually give rise to the well-known crop, pM = price of mineral fertilizer, pOr = price of organic fertilizer,
Marshallian inefficiency, in which both current production effect and investment are pW = price of the wood minus the logging and transportation cost, pF
provided below their first best level (Banerjee and Ghatak, 2004).
5 the price of the fruit, pP = price of the seedlings of the trees and its
In order to simplify notation of the theoretical model we denote mulch and
organic manure by the single variable XOr because both affect soil capital in a very
7
similar way. However, in the empirical part of the paper we distinguish between these Since the time perspective of owner-operated is independent of type of property
two organic fertilizers because farmers usually apply only one of them. right, we simply use the term owner in the theoretical section of the article to refer to
6
In a recent paper, Jacoby and Mansuri (2008) indicate that field trials in Pakistan both types of owners.
8
show that the marginal effects of manure on grain yields persist for at least three years Inherent to sharecropping is the question of sharing the risk between the landlord
following the initial application, while the productivity effects of mineral fertilizers are and the farmer. However, as we focus on the issue of different tenure regimes we use
essentially limited to the season of application. expected values and do not analyze the variation in crop yields.
 A. Abdulai et al. / Journal of Development Economics 96 (2011) 66–78 69

plantation, δ = degradation of the soil capital, and φ = discount rate. provided that the tenant or sharecropper has conserved or invested in
The previously introduced parameters ϑ, γ1, γ2 and p̄L also belong to the soil, so that soil capital exceeds a threshold value Stk at the end of
the vector ῡ. the initial lease arrangement, i.e., S(tk) ≥ Stk. The farm net benefits of a
Let the solution of problem (1) be given by tenant or sharecropper with no extended tenure arrangement over a

 planning horizon of 3k is given by
     = J  ðt ; t ; Sðt ÞÞ;
J L ðt Þ; XOr ðt Þ; XM ðt Þ; P ðt Þ; C ðt Þ; t0 ; tk ; Sðt0 Þ; υ 0 k 0
2  
J˜Te;Sh = ∑ J tik ; tði
 
where the superscript * indicates the evaluation of the variable along + 1Þk ; Sðtik Þ:
i=0
its optimal trajectory given the parameter values of t0, tk, S(t0) and ῡ.
Hence, J*(t0, tk, S(t0)) indicates the maximized discounted farm net while the farm net benefits of an extended tenure arrangement after
benefits aggregated over the time horizon of length tk − t0 given the the first period can be expressed as
initial soil capital of S(t0) = St0.
As indicated in the Introduction, observations by some authors ˜  
J˜Te;Sh = J t0 ; tk ; Sðt0 Þ; Sðtk Þ N Stk + J ðtk ; t3k ; Sðtk ÞÞ:
 
show that farmers may undertake land-improving measures such as
planting of trees in order to gain land rights, resulting in endogeneity
of land rights in an investment specification. We analyze this potential
Observation 2. The condition J˜Te;Sh b J̃˜Te;Sh indicates that it is optimal
endogeneity effect by examining the extent to which soil-improving  
investment choices, L(t), XOr(t), XM(t), P(t), C(t), may affect land tenure
for a tenant or sharecropper to conserve or invest in the soil in exchange
arrangements.9 For this purpose, we consider two cases: a) invest-
for an extended tenure arrangement of length 2k, whereas J˜Te; Sh N J̃˜Te; Sh
 
ment in land (land acquisition) and b) investment in soil capital (soil
indicates that it is optimal to continue with tenure arrangements of
improvement): Since the decision model (1) identifies farm returns,
length k.
the costs of fertilizer, young trees, and land, the farm net benefits
already take into account the costs and benefits associated with Observation 2 suggests that a farmer who maximizes farm net
different land tenure arrangements. Thus, it is possible to compare the benefits over a planning horizon of length 3k should conserve or
discounted stream of future farm net benefits of the different land invest in soil capital if the additional farm net benefits from improved
rights. Define n N 0 as the number of potential contracts and k as the soil are greater than the associated investment costs. In this case, soil
duration of each contract (in years). Hence, the discounted farm net conservation or soil improvement leads to a change in the conditions
benefits of a tenant (Te) or share cropper (Sh) aggregated over a time of the land tenure arrangement.10
n−1

 
Sh = ∑ J tik ; tði + 1Þk ; Sðtik Þ . Thus,
span of n k years, are given by: JTe; After analyzing the extent to which soil investment choices may
i=0 affect land tenure rights, we now examine the extent to which land
the discounted farm net benefits are maximized for each contract of
rights may affect soil investment choices. For this purpose, we
length k not taking into account the other n − 1 contracts. The
evaluate the first order conditions of the farmer's decision problem
different periods are connected through the stock variable since the
given in Eq. (1). To simplify notation, we suppress the argument t on
terminal value of period i becomes the initial value of the soil capital of
the variables as well as those of the costate variables and Lagrange
period i + 1. Consequently, a tenant or sharecropper maximizes the n-
multipliers to be introduced later, and define the current value
sequence of farm net benefits of k-years, but does not maximize over
Lagrangian, ℒ, by
the time horizon of n k. In contrast, an owner (O) maximizes the farm
net benefits over the entire planning horizon n k. Thus, the discounted ℒ = ðpf ðS; XM ; XOr Þ−pM XM −pOr XOr ÞL−pL ð⋅Þ + pW C + pF ϑ W
maximized farm net benefits over the entire planning horizon are
given by J*O = J*(t0, tnk, S(t0)). −pp P + μ1 L + μ2 ð1−αW−LÞ + ξ1 XM + ξ2 XOr + ξ3 P + ξ4 C

* * + λS ðhðXOr ÞL + yðW Þ−δf ðS; XM ; XOr ÞLÞ + λW ðg ðW Þ−C + P Þ;

Observation 1. The condition JTe, Sh b JO indicates that it is optimal
for a tenant or sharecropper to acquire land permanently, whereas JTe, ð2Þ
* *
Sh N JO indicates that it is not optimal.
where λS and λW are the corresponding costate variables, μ1 and μ2 are
Observation 1 points out that a tenant or sharecropper who Lagrange multipliers associated with the restrictions related to the
maximizes farm net benefits over a planning horizon of length n k availability of land, and ξ1 to ξ4 are Lagrange multipliers related to the
should acquire land on a permanent basis, if the costs of investing in non-negativity of the control variables. The first order conditions are
land and soil capital are lower than the sum of the additional farm net given by
returns resulting from the improved soil and the saved payments in
∂ℒ
cash or kind for the land-use. In this case, land improvement and land = pf −pM XM −pOr XOr −θγ1 pf + λS ðh−δf Þ + μ1 −μ2 = 0 ð3Þ
acquisition decisions occur simultaneously. Finally, land acquisition ∂L
may also be considered as a mean of securing a stream of future farm  
∂ℒ
net returns in order to avoid a situation where the tenure = pfXM −pM −ðθγ1 p + λS δÞfXM L + ξ1 = 0 ð4Þ
∂XM
arrangement is not renewed after m contracts, 0 b m b n, in which
case the analysis would have to consider the expected value of J*Te, Sh,   
∂ℒ ′
rather than its certainty value. = pfXOr −pOr −θγ1 pfXOr + λS h −δfXOr L + ξ2 = 0 ð5Þ
∂XOr
In the second case, we consider investment in soil capital (soil
improvement). In particular, we analyze the situation where tenants
or sharecroppers invest in soil capital during the first tenure period of
10
length k in order to achieve an extended tenure arrangement of length As pointed out by a reviewer, the current formulation allows the landlord to
2k instead of k in the subsequent period. We assume that the landlord extract the increase in soil capital as a result of a “hold-up problem”, i.e., the landlord
as the owner of the asset can withhold the benefits of an increase in the asset quality.
is willing to offer an extended duration of the tenure contract
Alternatively, one can think of different contract arrangements that allow the tenant to
internalize part of the increase in the asset quality, for instance, if the landlord
9
As we discuss in the Empirical analysis section of the paper, Jacoby and Mansuri commits to renegotiate the conditions of the contract at a certain point in time, while
(2008) argue that most models of agrarian contracts result in endogeneity of land the contract is in place. See the paper by Jacoby and Mansuri (2008) for a discussion of
tenure variables. the importance of imperfect commitment.
70 A. Abdulai et al. / Journal of Development Economics 96 (2011) 66–78

∂ℒ
= −pP + λW + ξ3 = 0 ð6Þ
A pfX ,with S(t) = S0
∂P M
pfX ,with S(t) > S0
M
pfX ,with S(t) < S0
M
∂ℒ
= ð1−θγ2 Þp′W + ξ4 −λW = 0 ð7Þ pM +λ Sδ fX (owner)
∂C M

pM +γ 1 pfX (sharecropper)
M
λ̇S = φλS −ððp−θγ1 p−λS δÞfS LÞ ð8Þ
pM (fixed-rent tenant)

′ ′
λ̇W = φλW −pF ϑ + αμ2 −λS y −λW g ðW Þ: ð9Þ
O
XM O
XM Sh
XM Sh
XM Te Te
XM XM
XM

If we assume that the initial soil stock S(0) = S0 is identical for all B
three tenure regimes, we can determine the optimal short-run pf Xor , S(t) = S0
behavior for all three types of tenure arrangements. However, the pf Xor , S(t) > S0
soil stock will not be identical for all three tenure regimes over time pf X , S(t) < S0
or
and we also need to account for individual changes in S(t). These
por+ λ S( −h′+δ fX ) > por (owner)
changes are significant in determining the optimal long-run behavior or

of the different types of farmers. The short-run and long-run pM +γ 1 pfXor (sharecropper)
behaviors of farmers will be discussed later.
pM(fixed-rent tenant)

3.1. Short-run behavior

por+ λ S( −h′+δ fX ) < por (owner)
or

For an interior solution, (ξ1 = 0), and pfXM(⋅)|S(t) = S0 the solution of

Eq. (4) is presented in Fig. 1A.11 Whereas owner-cultivated will O Sh Sh Te Te O Xor
X or X or X or X or X or Xor XorO
consider the shadow cost of the soil (λS), tenants or sharecroppers do
not consider these costs. Hence, fixed-rent tenants, sharecroppers and Fig. 1. A. The optimal amount of mineral fertilizer. B. The optimal amount of organic
fertilizer.
owners apply XTe Sh O
M , XM , and XM levels, respectively.

Observation 3. Given that the initial soil capital is identical for all
three types of farmers, Fig. 1A shows that tenants initially apply more
mineral fertilizer than sharecroppers and owner-cultivators. A direct 3.2. Long-run behavior
comparison between sharecroppers and owner-cultivators is not possible
in the present analysis. For the determination of the long-run behavior of the farmers, we
need to establish the optimal evolution of soil capital (stock variable).
The behavior of sharecroppers will depend on the value of γ1. In For this purpose, we distinguish between the situations where the
case the owner chooses γ1 = λSδ / p, owners and sharecroppers would initial soil capital is above the long-run soil capital and where it is
tend to apply the similar levels of mineral fertilizer. The optimal levels below. Although fixed-rent tenants and sharecroppers usually have
of organic fertilizer, which can be derived from an interior solution of short-term contracts, we include their behavior in the analysis by
Eq. (5) is presented in Fig. 1B. interpreting the sequence of their short-term behavior as their long-
run behavior. The principal results can be summarized in the
Observation 4. Given that the initial soil capital is identical for all
following observation.
three types of farmers, Fig. 1B shows that owners apply initially more
organic fertilizer than tenants, if the soil improvement effect of organic Observation 5. Provided that owners build up soil capital, Fig. 1A
fertilizer outweighs its degradation effect. Sharecroppers apply less and B shows that owners will tend to reduce the application rate of
organic fertilizer than tenants. mineral and organic fertilizer over time, whereas sharecroppers and
fixed-rent tenants increase the application of mineral and organic
Given that owners consider the shadow cost of the soil (λS), they fertilizer over time, provided that their soil capital decline.
apply more organic fertilizer than a tenant, provided that the soil
improvement effect of organic fertilizer, h′, is greater than the soil The demonstration of Observation 5 and additional analysis are
degradation effect (δfXOr) of cultivation. This situation is depicted in presented in Appendix A. Condition (6) suggests that it is optimal to
O
Fig. 1B by comparing XTe Or with X̃ Or . It is however significant to note plant young trees if their in-situ value, λW, is equal to their planting
that tenants may apply more organic fertilizers than owners under cost. Otherwise, ξ3 presents the difference between planting cost
specific conditions. Such a situation may arise, if the soil improvement and in-situ value of the young trees, and it is optimal to plant no
effect of organic fertilizer is lower than the soil degradation effect, trees. Cutting of trees results in C N 0, and therefore ξ4 = 0 in Eq. (7).
which compares XTe O
Or with XOr. Fig. 1B also indicates that sharecroppers Hence, for tenants and owners, the unit price of wood needs to be
apply less organic fertilizer, XSh Or, than fixed-rent tenants. A direct equal to the in-situ value of the standing trees, while an additional
comparison between sharecroppers and owners is however not γ 2p W has to be subtracted from the net price of wood for
possible.
 If the share that accrues to the landlord is equal to sharecroppers. If the net price minus θγ2PW is not equal to the in-
λS −h′
+ δ , owners and sharecroppers would apply similar levels situ value, the difference will be reflected in the value of ξ4. It is also
p fxOr
of organic fertilizer. evident in Eq. (7) that λW is equal to pW(1 − θγ2) for cases where
C N 0, indicating that λW is a constant, and hence λ̇W = 0. This
11
Since the cross derivative of fXMXOr is zero, we can graph pfXM independently of the
condition holds at the steady state equilibrium by definition.
tenure arrangement, although XM and XOr vary with the tenure arrangement. However, Moreover, the condition λ̇W = 0 holds outside the steady state
pfXM varies with S, since fXMS and fXOrS decrease with an increases in S. equilibrium when farmers cut trees. Hence, the following discussion
 A. Abdulai et al. / Journal of Development Economics 96 (2011) 66–78 71

holds for the two described situations. In this case, utilizing the differences in μ2 are relatively small and do not alter the ranking of the
definition of λW in Eq. (7), (9) can be written as: optimal W for the different tenure regimes. The empirical analysis
  undertaken with primary data addresses the situation where the
′ ′
0 = φ−g ðW Þ ð1−θγ2 ÞpW −pF ϑ + αμ2 −λS y ðW Þ: ð10Þ previous assumption is not applicable.

Observation 6. If the opportunity costs of land are initially zero

where land is in abundance, the opportunity cost of land tends to be (land abundance) and increase thereafter, for instance as a result of an
zero, i.e., μ2 = 0. Given that both tenants and sharecroppers do not increase in population pressure; owners and tenants grow fewer trees if
consider the shadow cost of the soil, Eq. (10) reduces to they are used exclusively for wood production. However, if the
  opportunity costs of land are strictly positive from the beginning, the
′
0 = φ−g ðW Þ ð1−θγ2 ÞpW −pF ϑ: ð11Þ influence of land tenure rights on tree planting can only be derived
empirically.
Let us first consider the case where the trees only produce wood
without fruits, i.e. ϑ = 0. In this situation, Eq. (11) holds, if we The demonstration of Observation 6 is presented in Appendix A. The
choose W such that g′(W) is equal to φ, in which case the marginal foregoing analysis has considered the case where trees are cut, that is,
growth rate of the biomass will be equal to the discount rate. This C N 0. In the specific case where C = 0 we know from the relationship
case is depicted in Fig. 2 for W = W*, a result that is standard in Ẇ = g ′ ðW Þ−C + P, that W is defined by dW dt
= g ′ ðW Þ = ∫dW =
t
natural resource economics for common property resources, since ∫g ′ ðW Þdt, in which case W ðt Þ = WO + ∫0 g ′ ðW Þdt. Trees grow during
the optimal stock is on the left side of the stock that supports the this phase without being cut.
maximum sustainable yield, WMSY. In the situation where trees
produce wood and fruit, ϑ N 0, Eq. (11) can only hold if the term 4. Empirical analysis
(φ − g′(W) is strictly positive. In other words, the term g′(W) has to
decrease in comparison with the situation where trees do not In this section, we employ plot-level data to examine the
produce any fruit. Consequently, as it can be seen from Fig. 2, the relationship between land tenure arrangements and investments in
optimal amount of W has to increase and will be situated to the left productivity-enhancing measures, to complement the theoretical
of W*. Hence, if trees produce fruit, farmers find it optimal to analysis presented in the previous section.12 We also investigate how
increase the “number of trees”. land tenure arrangements affect farm productivity.
However, if agricultural land is scarce, the opportunity cost of land
is given by μ2 = pf − pMXM − pOrXOr − θγ1pf + λS(h − δf), according to 4.1. Econometric specification
Eq. (3). Hence, the optimal W for owners is given by
The first order conditions (4)–(9) imply that farmers invest in
!
′ p ϑ αμ2 λS y
′ land-improving or conservation measures if it leads to an increase in
0= φ−g ðW Þ− F + − pW
pW pw pW the expected farm net benefit aggregated over the planning horizon.
!!! However, the expected farm net benefit is not observable, since it is
′
α pF ϑ y
= φ−g ′ ðW Þ + − + pf −pM XM −pOr XOr + λS h−δf − pW subjective. What is observed is the decision to invest or not to invest,
pW α α
i.e. the planting of trees, application of mineral fertilizer, as well as
ð12Þ organic fertilizer such as mulch and organic manure. The empirical
analysis focuses on the factors that influence the likelihood of farmers
and for sharecroppers and tenants by
engaging in these investments. In line with the maximization problem
! outlined in Eq. (1), farmers invest in soil-improving and natural
′ αμ2 −pF ϑ
0= φ−g ðW Þ +
2 ð1−θγ2 ÞpW resource management measures, if it augments the farm net benefit,
1−θγ2 pW
 !
i.e. ∂ J / ∂ P, ∂ J / ∂ XM, ∂ J / ∂ XOr N 0.
α p ϑ ð13Þ Unfortunately, changes in J are not observable, but can be
= φ−g ′ ðW Þ +
2 − F + pf −pM XM −pOr XOr −θγ1 pf
1−θγ2 pW α expressed as a function of observable elements. Let us define the
ð1−θγ2 ÞpW : underlying latent propensity variable for investment on plot j, owned
by farmer l, for the soil-improving and natural resource management
It needs to be noted that the optimal W cannot be unambiguously strategy m as Ĵ jlm . The underlying propensities can then be related to
determined from Eqs. (12) and (13), since the term μ2 is not identical the plot's observed characteristics and farmer related variables, Zjlm, as
for the different tenure regimes. Hence, the graphical solution of well as land tenure arrangements, Rjlm and unobserved characteristics,
Eqs. (12) and (13) can only be obtained under the assumption that the εjlm, in the following latent variable model:

Ĵ jlm = Zjlm βm + Rjlm γm + εjlm ðm = Trees; Fertilizer; Mulch; ManureÞ:

ð14Þ

Denoting tree planting, mineral fertilizer application, mulching

and manure application as P, XM, XMu Ma
Or , and XOr , respectively, Eq. (14)

12
As noted by a reviewer, the theoretical analysis provides general results in terms of
relative values, of course, at the cost of some simplification of the complex real world.
The empirical analysis complements the theoretical analysis with plot-level data of
farm households. However, the results presented are in terms of probabilities due to
the fact the changes in the farm net benefits are not observable. Hence, the empirical
study should be considered as complementary analysis rather than an exact
Fig. 2. The optimal “number of trees” in the presence of agricultural production. specification of the theoretical model.
72 A. Abdulai et al. / Journal of Development Economics 96 (2011) 66–78

can be transformed into a binary probit equation for participation for where Rjlm is a vector of the observed tenure arrangement variables and
each investment option under the following mapping from the latent Ujlm is a vector of the residual terms from Eq. (16). The probit estimates
variable to its observed realization: of β2 in Eq. (17) are consistent (Blundell and Smith, 1989; Wooldridge,
8 2002).16 A significant feature of the approach is that the usual probit t-
>
< 1 if Ĵ jlm N 0;
> statistics on β3 are valid tests of the null hypotheses that the variables
Ĵˆjlm = 0 if Ĵ jlm ≤0:   ð15Þ are exogenous, i.e., H0 : β3 = 0. However, if β3 ≠ 0, then the probit
>
>
: Mu Ma
m = P; XM ; XOr ; XOr : standard errors and test statistics are not strictly valid, and we would
have only estimated β1 and β2 up to scale (Wooldridge, 2002). As
pointed out by Wooldridge (2002), under H0 : β3 = 0, ξjlm = μjlm, and
Let us assume that εjlm(m = P, XM, XMu Ma
Or , XOr ) jointly follow a
hence the distribution of ξjlm plays no role under the null hypothesis.
multivariate normal distribution with mean zero and variance 1,
Therefore the test of exogeneity is valid without assuming normality or
and the covariance matrix Σ.13 This can be expressed as
homoskedasticity of εjlm. However, when Rjlm and εjlm are correlated,
 ′ normality of Ujlm is crucial. As noted by Brasselle et al. (2002), a joint
εP ; εXM ; εX Mu ; εX Ma eMVNð0; ΣÞ: Wald test can also be performed on the vector of β3's to examine the
0 0

exogeneity of tenure arrangements as a whole.

Maximum likelihood method can then be employed to estimate While the specification in Eq. (17) controls for farmer and plot
the parameters and the four correlations of the error terms (Greene, characteristics, it does not control for family-level unobservables that
2008). However, because the probabilities that enter the likelihood could be correlated with both tenure status and the decision to adopt
are functions of high dimensional multivariate normal distributions, land-improving measures.17 To deal with this concern, we allowed for
they are simulated using GHK algorithm (Greene, 2008, p. 582). Some household fixed effects in Eq. (17) to yield the following specification
studies on investment in soil-improving and natural resource
management measures have employed single-equation techniques,
Ĵ jlm = λlm + β1 Yjlm + β2 Rjlm + β3 Ujlm + ζjlm ð18Þ
with the assumption that εjlm independently follows univariate
distributions.14 However, because of the substitutability or comple-
mentarity between these investment options, and the fact that the where λlm is the household-specific intercept and represents the
plots in the sample are similar across equations, it is most likely that intrinsic propensity (based on variables unobserved by the researcher)
the error terms of these equations will be correlated. of farmer l for activity m; Y is vector of plot-level variables. Thus, in
As indicated earlier, land tenure rights may be influenced by estimating the household fixed-effects model, all household level and
investment decisions, resulting in endogeneity of the land tenure district level variables drop out of the regression. A linear probability
arrangement variables in the investment specification. Jacoby and model is employed in the first-stage estimation of the four tenure rights
Mansuri (2008) also point out that most models of agrarian contracts variables.18 Specifically, three tenure arrangement variables are
imply a correlation between contractual choice and unobserved estimated, with owner-operated without transfer rights being used as
cultivator characteristics, resulting in endogeneity of land tenure the default variable.
arrangements. A properly specified two-stage instrumental variable
approach will produce consistent estimates of the parameters in 4.2. Identification strategy
Eq. (14). The first-stage equation specifies land rights as a function of
exogenous variables, including those in Eq. (14) and others that affect To ensure identification in the estimation of the investment
land tenure arrangements, specification, some of the variables included in the first-stage
estimation of tenure rights are excluded from the multivariate probit
Rjlm = α0 + Zjlm α1 + Vjlm α2 + ξjlm ; ð16Þ estimation. As suggested by Jacoby and Mansuri (2008), a suitable
identification strategy is to employ a variable that strongly influences
where Vjlm is a vector of instrumental variables that is correlated with contractual choice but is orthogonal to unobserved plot character-
land tenure arrangements but uncorrelated with εjlm, the residual in istics. Distance of the plot from the landlord's home is a suitable
Eq. (14), and is therefore excluded from Eq. (14). The predicted values variable, given that plots that are located further away from home are
from Eq. (16) are then used in the second stage estimation of Eq. (14). more costly to self-cultivate. Now to serve as an excluded instrument,
However, when the dependent variable is discrete as in the present a question that needs to be addressed is whether distance to the
study, the usual two-stage approach described previously will not be landlord is related to plot quality.19 To examine the relationship
able to address the endogeneity problem.15 Wooldridge (2002) between proximity of plot from landlord and plot quality, we
argues that the most useful two-step approach to examine endo- estimated reduced-form regressions of soil quality, controlling for
geneity in a probit model is the Two-Stage Conditional Maximum plot characteristics and district dummies (Jacoby and Mansuri, 2008).
Likelihood (2SCML) proposed by Rivers and Vuong (1988). Proximity is defined by whether landlord resides in the village the
Rather than using the predicted values from the first-stage linear plot is located, and the distance of plot from the landlord' residence.
probability regression, the 2SCML approach involves specifying the The results, which are not presented for the sake of brevity show that
investment equation as proximity does not significantly influence soil quality.
 
Mu Ma
Ĵ jlm = β0 + β1 Zjlm + β2 Rjlm + β3 Ujlm + μjlm m = P; XM ; XOr ; XOr ;
16
Rivers and Vuong (1988) point out that the usual probit standard errors and test
ð17Þ statistics are not strictly valid if the null hypothesis of exogeneity of the variable is
rejected. In such a case, they suggest the use of an M-estimator to derive the
asymptotic variance of the two-step estimator.
17
We are grateful to an anonymous reviewer for drawing our attention to this point
13
As pointed out by Greene (2008), the magnitude of the variance of the disturbance and suggesting the use of household fixed-effects analysis.
18
term cannot be identified for each probit equation, as such the variance has normally Brasselle et al. (2002) also employed the 2SMCL in their study on Burkina Faso,
been assumed as 1. using a linear probability model in the first stage, while Besley (1995) employed the
14
For example, Marenya and Barrett (2007) employed single probit models for the linear probability model to estimate the investment specification in his study.
19
investment options in their study on Western Kenya. As pointed out by Jacoby and Mansuri (2008), landlords may purchase poor
15
The non-linearity of the probit model will result in estimates of standard errors quality land far away from their homes with the intention of leasing it out. This is not
that are downward-biased and coefficients that are not normally distributed likely to be the case in this sample, since land is mainly acquired through inheritance,
(Wooldridge, 2002). with only about 1% acquired through purchases.
 A. Abdulai et al. / Journal of Development Economics 96 (2011) 66–78 73

The variables that are excluded from the second stage regression of Information on household characteristics, such as number of years of
investment decisions for owner-cultivated plots are mode of schooling and age of farmer, value of livestock owned by household,
acquisition of the plot, a dummy variable indicating whether and value of farm implements were included.
cultivator resides in the village where the plot is located or not, and Differences across plots in terms of quality and location also affect
the distance of the plot from the cultivator's home. The variables used the suitability of the plots for various investments. Information on
as instruments for fixed-rent tenants and sharecroppers are a dummy plot characteristics was therefore collected to address this issue. The
variable indicating whether the landlord resides in village where the plot-level characteristics gathered include plot size, distance of plot
plot is located or not, and the distance of the plot from the landlord's from home, gender of farmer cultivating plot, whether the cultivator
residence.20 The validity of the exclusion restrictions are tested with of the plot lived in the village where the plot is located, plot fertility,
the approach suggested by Lee (1992), who presents an overidenti- as well as the crops grown on the plot. The descriptive statistics of
fication test statistics, distributed as χ2 with degrees of freedom equal the variables used in the analysis are provided in Table 1. The
to the number of excluded instruments.21 The test involves estimating incidence of investment is measured by dummy variables that take
an alternative version of Eq. (17) that includes the instruments22 on the value of one when a particular investment was undertaken on
 a given plot and zero otherwise. Four variables are employed in the
′ ′ ′ ′ ′ Mu Ma
Ĵ jlm = β0 + β 1 Zjlm + β 2 Rjlm + β 3 Ujlm + β 4 I + μjlm m = P; XM ; XOr ; XOr Þ: study to examine tenure security. As mentioned previously, these
ð19Þ include owner-operated with full rights, owner-operated with
restricted rights, fixed-rent cultivation and sharecropping contract.24
The insignificance of β′4 then provides direct evidence that the All these variables are measured with dummy variables. Total farm
instruments can be excluded from Eq. (17). As indicated in the output value, measured in Ghanaian Cedis, includes revenues from
theoretical analysis, the magnitude of the influence of different tenure both crops and trees.
arrangements on investment decisions of farmers, as well as the sign
of the influence on some investment decisions cannot be determined 6. Estimation results
a priori, and therefore needs to be determined empirically, which will
be undertaken later. 6.1. Investment specification

5. Data and definition of variables Although our primary interest is on the impact of tenure
arrangements on investments, we first estimated the determinants
The data used in the analysis were collected during January and of land tenure arrangements to account for potential endogeneity
October 2003 in six villages in two districts – Techiman and Nkoranza – between the two variables. The results of the first-stage regression are
in the Brong Ahafo region of Ghana. A stratified random sample of 246 reported in Table A1 in Appendix A. In all three specifications, the p-
farm households with 560 plots was selected from four villages in values of the F-statistics for the instruments used indicate that they
Techiman District and two villages in Nkoranza District, with several play an important role in land tenure arrangements. Also reported in
households cultivating multiple plots with different land tenure the table are estimates that account for household fixed effects. The
arrangements. The locations sampled in Techiman include Twimea– estimates show that the signs of the coefficients are similar to the
Nkwanta, Aworopata, Woraso and Nkwaeso. In Nkoranza District, estimates without household fixed effects, although the magnitudes
Dromankese and Ayerede were sampled. of the coefficients vary slightly.
The sample was taken to ensure representation of the various land The empirical results for the investment specifications are
tenure arrangements in the area. The survey requested each presented in Table 2. It can be observed from the results that all
household to report its land tenure arrangements on each plot it estimated correlation coefficients are positive and significantly
cultivated. Specifically, it consisted of 202 owner-operated with different from zero at the 1% level of significance, indicating that
transfer rights plots, 104 owner-operated without transfer rights unobserved variables involved in each investment option are
plots, 159 plots under fixed-rent contracts, and 95 plots under significantly positively related, and confirms that it is more efficient
sharecropping contracts. A total of 194 households were identified to model investment decisions jointly rather than separately. The χ2
with multiple plots. Farmers were also asked about investments they statistics for the validity tests of the overidentifying restrictions
had undertaken in the past five years to improve the land they were presented in Table 2 fail to reject the exclusion restriction that the
cultivating. The investments include tree planting, mulching, manure instruments affect investment only via land tenure arrangements.
and mineral fertilizer.23 Noteworthy is the fact that all the variables representing the
Land purchases are very rare in this area, where they account for residuals derived from the first-stage regressions for tenure arrange-
just 1% of all parcel acquisitions in the sample. On the other hand, ments are not statistically significant at conventional levels, indicating
inheritance through the family and borrowing are more common in no simultaneity bias and that the coefficients have been consistently
the area. All land acquisitions for owner-operated without transfer estimated (Brasselle et al., 2002; Wooldridge, 2002). Also shown in
rights were through borrowing and gifts, while inheritance and gifts the table are the χ2 statistics for the joint Wald tests on the vector of
were the main channels of land acquisition for owner-operated with these residuals from the first-stage estimations. These values also
transfer rights. Gifts were mainly from family members and friends. reveal that for each investment equation, the null hypothesis that the
residuals are jointly equal to zero could not be rejected, again
20
It is important to note that besides raising the cost of self-cultivation, greater confirming the results of the individual t-statistics.
distance between landowner and plot may also increase the cost of monitoring the
The variable representing owner-operated with full rights is
plot, if leased out, resulting in potentially biased estimates (Jacoby and Mansuri, 2009).
However, sharecroppers, who are likely to be monitored, are normally not supervised
positive and significantly different from zero in all four investment
in the study area. Hence, this problem does not arise in our estimation strategy. options, suggesting that relative to owner-cultivated without transfer
21
Davidson and Mackinnon (1993) explain that this statistic tests the joint rights, the owner-cultivated with transfer rights are more likely to
hypothesis that the excluded instruments are not inappropriately excluded and are invest in trees, organic and mineral fertilizers, confirming the
uncorrelated with the error term in the investment specification.
22 importance of secured property rights. Consistent with the theoretical
This involves specifying individual probit models for the various investment
activities and then employing the χ2 test for the validity of the instruments.
23 24
Very few farmers constructed ditches on their plots. Given the insignificant Although the typical fixed-rent tenure duration for the sample was 2 years, some
number of farmers that were engaged in this investment, we deleted it from the of the fixed-rent contracts went as high as to 4 years, while some of the 2 year contacts
analysis. had been renewed a number of times.
74 A. Abdulai et al. / Journal of Development Economics 96 (2011) 66–78

Table 1 Table 2
Descriptive statistics of variables used in the regression models. Multivariate probit regressions of investment in land improvement measures.

Variable Definition of variables Mean S.D. Variable Trees Mulch Fertilizer Manure

Dependent variables CONSTANT − 0.518⁎⁎⁎ − 0.296⁎ − 0.408⁎⁎ − 0.482⁎

TREES 1 if farmer plants trees, 0 otherwise 0.43 0.50 (3.26) (1.67) (2.29) (1.78)
FERT 1 if farmer applies fertilizer, 0 otherwise 0.42 0.49 OWNER 0.718⁎⁎⁎ 0.663⁎⁎⁎ 0.229⁎⁎ 0.335⁎⁎
MULCH 1 if farmer applies mulch, 0 otherwise 0.35 0.48 (4.19) (2.83) (2.07) (2.32)
MANURE 1 if farmer applies organic manure, 0 otherwise 0.14 0.34 SHARECROP 0.241 − 0.339⁎⁎ − 0.094 − 0.381⁎
YIELD Total output value per acre (1000 Cedis) 96.17 28.20 (1.28) (2.08) (0.91) (1.96)
FIXRENT − 0.365⁎⁎⁎ − 0.181⁎⁎ 0.277⁎ − 0.159⁎
Tenure variables (3.24) (2.36) (1.93) (1.88)
OWNER 1 if land is under own-operated with full rights 0.36 0.48 PLTSIZE 0.034⁎⁎ − 0.025 0.103⁎⁎⁎ − 0.115⁎⁎
FIXRENT 1 if land is under fixed-rent contract 0.28 0.45 (2.19) (0.68) (2.66) (2.14)
SHARECROP 1 if land is under sharecropping contract 0.17 0.38 PLOTFERT 0.281 0.573⁎⁎⁎ − 0.148 0.625⁎⁎⁎
OTHER 1 if land is under owner-operated without rights 0.19 0.26 (1.54) (3.08) (0.74) (2.81)
PLTYRS 0.082 0.196 0.106 0.073
Household characteristics (1.16) (0.88) (1.27) (0.92)
AGE Age of farmer (years) 49.98 13.67 FEMALE − 0.481 − 0.9764⁎⁎ 0.326 − 0.311
EDUCN Years of formal education of farmer 3.76 4.88 (1.46) (2.10) (1.04) (0.64)
LIVEST Value of livestock wealth (¢ × 10− 6) 11.20 26.11 ETHNIC 0.272 0.064 0.177 0.008
IMPLTS Number of implements owned by farmer 13.47 8.78 (1.41) (1.09) (1.56) (0.93)
EXTEN 1 if farmer received extension visit, 0 otherwise 0.38 0.49 AGE − 0.052⁎⁎⁎ 0.018 0.005 − 0.027⁎
(2.85) (0.86) (0.56) (1.69)
Plot characteristics EDUCN 0.092⁎⁎⁎ 0.056⁎⁎⁎ 0.026⁎ 0.014⁎
FEMALE 1 for female held plot, 0 otherwise 0.52 0.21 (4.34) (2.79) (1.87) (1.79)
PCREDIT 1 if household has access to credit, 0 otherwise 0.43 0.32 HHSIZE 0.017 − 0.028 − 0.025 0.021⁎
PLTDIST Distance of plot from landowner's residence (km) 2.33 1.91 (0.64) (1.20) (1.12) (1.75)
PLTLOC 1 if plot outside village of landowner, 0 otherwise 0.19 0.37 LIVEST − 0.006 − 0.002 0.012⁎ 0.032⁎⁎⁎
PLTSIZE Plot size in acres 2.94 2.03 (1.11) (0.03) (1.82) (2.73)
PLTYEARS Number of years the plot has been under use 16.58 9.46 IMPLTS 0.536⁎⁎⁎ 0.018 0.565⁎ 0.318
PLOTFERT 1 if plot is on fertile land, 0 otherwise 0.14 0.35 (3.63) (0.71) (1.82) (1.56)
RESOWNER 0.172 0.014 0.056 0.096
Crops (1.29) (1.37) (1.19) (1.53)
PLANTAIN If farmer cultivates plantain on plot, 0 otherwise 0.09 0.29 RESFIXED 0.034 0.016 0.024 0.043
CASSAVA If farmer cultivates cassava on plot, 0 otherwise 0.06 0.24 (1.06) (1.41) (1.32) (1.09)
VBEANS If farmer cultivates beans on plot, 0 otherwise 0.46 0.49 RESSHARE 0.134 0.082 0.064 0.116
OTHERS If farmer cultivates other crops, 0 otherwise 0.39 0.28 (1.08) (1.22) (1.41) (1.43)
Village dummy variables Yes Yes Yes Yes
Location dummies χ2-statistic for joint 1.14 0.87 0.91 1.02
TWIMEA 1 if farmer resides at Twimea–Nkwanta 0.22 0.41 significance of residuals [0.38] [0.46] [0.52] [0.39]
AWOROPAT 1 if farmer resides at Aworopata 0.13 0.34 χ2-statistic for 0.73 0.44 0.62 0.67
WORASO 1 if farmer resides at Woraso 0.18 0.38 overidentification (0.41) (0.33) (0.48) (0.51)
AYEREDE 1 if farmer resides at Ayerede 0.27 0.44 Cross-equation correlations
DROMA 1 if farmer resides at Dromankese 0.08 0.28 ρTF 0.256⁎⁎ ρTM 0.317⁎⁎⁎ ρTO 0.412⁎⁎⁎
ρFM 0.218⁎⁎⁎ ρMO 0.336⁎⁎⁎ ρFO 0.225⁎⁎⁎
Note: Exchange rate: US $1 = ¢8500 in 2003. ¢ = Ghanaian Cedis.
Mc Fadden R2 0.267

Note: Absolute t-values in parentheses. RESOWNER, RESFIXED and RESSHARE denote

residuals from the first-stage regressions for owner-operated with full rights, fixed-rent
and sharecropping contracts, respectively.
analysis, the variable for fixed-rent is negative and significant for
⁎ Significant at 10%.
trees, mulch and organic manure, but positive and significant for ⁎⁎ Significant at 5%.
mineral fertilizer. This indicates that relative to owner-cultivated ⁎⁎⁎ Significant at 1%.
without transfer rights, plots on fixed-rent contracts are less likely to
attract investment in trees, mulch and organic manure, but are more
likely to attract investment in mineral fertilizer for short-term rights. As explained earlier, farmers on fixed-rent contracts normally
benefits. try to maximize net benefits from the plots within a very short time,
Although the variables for sharecropping are negative, only those which usually includes applying relatively high levels of mineral
for mulch and manure are significantly different from zero, also fertilizer.25 The variable for sharecropping is negative in three of the
indicating that farmers under sharecropping are less likely to invest in specifications, but statistically significant only for manure. The
these activities, relative to owner-cultivated. The evidence clearly findings here clearly show that owner-operated with full rights are
reveals that mulch and organic manure are more likely to be used on more likely to invest in these activities than sharecroppers. The
plots that are owner-cultivated with full rights and owner-cultivated positive and significant impact of secured tenure rights on investment
without full rights than leased plots with the same characteristics. is in line with the results reported by Besley (1995) and Goldstein and
Trees are also more likely to be planted on plots that are owner- Udry (2008) for Ghana. In particular, the finding that farmers invest
cultivated with full rights and owner-cultivated without full rights
than on leased plots. The finding that owner-cultivated plots are more 25
To examine the effect of tenure duration (number of years current tenant has
likely to apply organic fertilizers than both sharecroppers and fixed-
cultivated the plot) on investments, we re-estimated the specifications in Table 2, and
rent tenants suggests that the short durations of the contracts that in the spirit of Jacoby and Mansuri (2008), included the log of tenancy duration in the
tend to reduce the benefits obtained by tenants from the application specifications, with zero for owned plots. The results revealed positive and significant
of manure and mulch may be serving as a disincentive for them to coefficients for mulch, trees and organic manure, suggesting that farmers with longer
invest in these measures. Moreover, it may grow stronger the closer contract durations are more likely to invest in soil-improving measures. Thus, the
longer the tenant stays on a leased plot, the higher the probability of his increased
the contract is to its expiration date. investment in measures that improve soil quality over an extended period, compared
The empirical results also indicate that fixed-rent tenants are more to short durations contracts that only encourage investment in measures that increase
likely to invest in mineral fertilizer than owner-cultivated without full crop productivity in a planting season.
 A. Abdulai et al. / Journal of Development Economics 96 (2011) 66–78 75

less in organic fertilizers (mulch and manure) in plots that are leased Table 3
than in those that are owned are consistent with the recent findings Multivariate probit regression of investment in land improvement measures with
household fixed effects.
by Jacoby and Mansuri (2008).
Trees are more likely to be planted by farmers with higher Trees Mulch Fertilizer Manure
education, more assets, and larger plot sizes. In particular, education CONSTANT − 0.432⁎⁎⁎ − 0.220⁎ 0.336 − 0.568
appears to have a positive and significant impact on all four (3.17) (1.81) (1.47) (1.39)
investment options, a finding that is in line with the human capital OWNER 0.616⁎⁎ 0.541⁎⁎ 0.165⁎ 0.329⁎⁎
(3.02) (2.27) (1.98) (2.19)
theory. Age exerts a negative and significant effect on both tree
SHARECROP 0.173 − 0.305⁎ − 0.076 − 0.268⁎⁎
planting and the application of manure. Thus, controlling for tenure (1.36) (1.97) (1.63) (2.15)
arrangements and other farmer's and plot-level characteristics older FIXRENT − 0.221⁎⁎⁎ − 0.218⁎ 0.234⁎ − 0.184⁎
farmers appear to be less likely to invest in trees and manure. This is (2.68) (1.82) (1.78) (1.92)
probably because younger farmers cultivate the land for a longer time, FEMALE − 0.217⁎⁎ − 0.53⁎ 0.036 0.114
(2.33) (1.81) (0.93) (1.07)
and as such are in a better position to benefit from the returns from
PLTSIZE 0.161⁎⁎ 0.075 0.118⁎⁎ − 0.066
investments in soil-improving measures even in the distant future. In (2.24) (0.91) (2.35) (1.29)
particular, if farmers are not credit constrained and take future PLOTFERT 0.043 0.528⁎ 0.012 0.137⁎⁎
generations into account, younger farmers will be more likely to (1.26) (1.83) (1.14) (2.16)
PLTYRS 0.012 0.031 0.019 0.024
invest in conservation measures than older ones. Almost all the village
(1.02) (0.98) (1.47) (1.35)
dummies are significantly different from zero, indicating significant RESOWNER 0.137 0.015 0.118 0.046
cluster effects, and probably revealing agro-climatic variation and (1.43) (1.19) (1.28) (1.06)
access to infrastructure.26 As noted by Besley (1995), they could also RESFIXED 0.019 0.027 0.014 0.052
be representing village-level variation in tenure arrangements. (1.22) (1.44) (1.38) (1.29)
RESSHARE 0.135 0.077 0.104 0.137
The estimates with household fixed effects are presented in
(1.19) (1.38) (1.12) (1.04)
Table 3. The results indicate that the magnitudes of the coefficients χ2-statistic for joint 0.86 0.68 1.12 0.94
representing the tenure arrangement variables are slightly lower than Sig. of residuals [0.41] [0.59] [0.36] [0.43]
2
those without household fixed effects. However, the positive impact χ -statistic for 0.76 0.57 0.65 0.84
Overidentification [0.55] [0.48] [0.47] [0.52]
of the secure tenure rights on investment in the productivity-
Cross-equation correlations
enhancing measures, as well as the negative effects of shared- ρTF 0.344⁎⁎
cropping and fixed-rent tenancy on applying organic fertilizer remain. ρFM 0.216⁎⁎⁎
This suggests that unobserved household effects are not responsible ρTM 0.279⁎⁎⁎
for the positive and significant impact of secured property rights on ρTO 0.326⁎⁎⁎
ρFO 0.238⁎⁎⁎
investment decisions.
ρMO 0.315⁎⁎⁎
Given that the coefficients presented in Tables 2 and 3 indicate the Mc Fadden R2 0.252
impact of the explanatory variables on the probability of each choice
Note: Absolute t-values in parentheses and p-values in squared brackets. RESOWNER,
but not the marginal effects, we compute the marginal contributions RESFIXED and RESSHARE are as defined in Table 2.
of the explanatory variables on the probability of investing in ⁎ Significant at 10%.
productivity-enhancing measures. We are particularly interested in ⁎⁎ Significant at 5%.
⁎⁎⁎ Significant at 1%.
the marginal effects of the land tenure arrangement variables. We
employ the estimates from Table 3 which account for household fixed
effects. The marginal effects are evaluated at the means of the instrumental variable analysis are presented in Table 5. Given the
explanatory variables. The standard errors of the marginal effects are significant diversity of crops and intercrops on the plots, we employed
estimated using the DELTA method (Greene, 2008). All the significant value of crop output per acre as the dependent variable (Place and
marginal effects have the expected signs. Looking at the tenure Hazell, 1993; Place and Otsuka, 2002). Separate analysis for each
arrangement variables, the evidence shows that being an owner- cropping pattern was not undertaken because of the relatively small
cultivator increases the probability of investing in soil-improving sample sizes that arise from the data set. Dummy variables for
measures between 11% and 49%. On the other hand being a fixed-rent cropping patterns were however introduced in the regression to
tenant tends to decrease the probability of investing in trees by 16%, capture the effects of the individual crops.
mulch by 24%, and manure by 12%. However, being a fixed-rent tenant Given the potential endogeneity of the access to credit variable, it
increases the probability of investing in mineral fertilizer by 16%, was instrumented by first estimating a probit model of determinants
which is even higher than the 11% for an owner-cultivator. of access to credit and then using the predicted values in the
Furthermore, being a sharecropper reduces the probability of productivity estimation. This is because in some cases, land or a crop
investing in both manure and mulch by 27% and 32%, respectively, itself can be used as collateral to obtain credit. The results from this
which are much higher than the reduction in probability of investing first-stage regression are not presented for the sake of brevity, but are
by fixed-rent tenants. These findings are consistent with the notion available from the authors upon request. The estimates in Table 5
that secured rights matter for investment in productivity-enhancing indicate a positive and statistically significant effect of the owner-
measures (Table 4). operated with rights variable, suggesting that ownership of land
results in higher output. The results actually reinforce the finding that
6.2. Tenure arrangement and farm productivity secured tenured rights facilitate investments in land-improving
measures or yield-enhancing inputs. This finding is consistent with
As pointed out by Jacoby and Mansuri (2008), it is worth asking results reported by Banerjee et al. (2002), who found a positive
how yields would be affected by land ownership, given the results impact of tenure reform on agricultural productivity for West Bengal
presented previously. We investigate the impact of tenure arrange- in India. The results are also consistent with the findings by Goldstein
ments on plot-level productivity in this section. The results from this and Udry (2008) for Ghana, who showed that a great deal of potential
output is lost in the study area because land tenure is insecure.
26 However, the findings contrast with those reported by Place and
The joint test of the null hypothesis that all district effects are equal using a
likelihood ratio test gives a sample chi-squared value of 72.48 and a critical value at Hazell (1993) and Place and Otsuka (2002), who found no significant
the 1% level of 16.8. relationship between tenure and crop productivity in their studies.
76 A. Abdulai et al. / Journal of Development Economics 96 (2011) 66–78

Table 4 7. Conclusion
Marginal effects on the marginal probability of investment (in %).

Trees Mulch Fertilizer Manure In this article, we developed a framework to examine the
relationship between different land tenure arrangements and house-
OWNER 0.3946 0.3853 0.1135 0.4877
(0.0817) (0.0938) (0.0421) (0.1029) holds' investment in land-improving and conservation measures in
SHARECROP 0.1108 − 0.3190 − 0.0523 − 0.2756 the Brong Ahafo region of Ghana. The land tenure arrangements
(0.0721) (0.0992) (0.0214) (0.0808) considered include owner-operated with full property rights, owner-
FIXRENT − 0.1609 − 0.2375 0.1579 − 0.1179
operated with restricted rights, fixed-rent and sharecropping con-
(0.0386) (0.0938) (0.0362) (0.0392)
FEMALE − 0.1190 − 0.0761 + 0.1476 0.0730 tracts. We employed variations in tenure arrangements between
(0.0613) (0.0396) (0.0548) (0.0317) different plots to estimate plot-level regressions relating tenure
PLTFERT 0.0275 0.0735 0.0770 0.0878 arrangement to investment in tree planting, mulch, manure as well as
(0.0108) (0.0313) (0.0269) (0.0365) mineral fertilizer application. We also examined the relationship
PLTSIZE 0.1031 + 0.1089 0.0812 − 0.0443
between land tenure arrangements and farm productivity.
(0.0792) (0.0421) (0.0367) (0.0219)
PLTYRS 0.0077 0.0182 0.0179 0.0152 The empirical results are consistent with our theoretical findings
(0.0015) (0.0098) (0.0086) (0.0084) and show that secured land rights tend to facilitate investment in soil-
Notes: Standard errors of the estimated marginal effects are presented in parentheses. improving and natural resource management practices. In particular,
farmers who owned land with secured tenure were more likely to
invest in tree planting, mulch, manure, but not in mineral fertilizer.
The fixed-rent variable also showed a positive sign, but is not Farmers on fixed-rent and sharecropping contracts were found to be
significantly different from zero at conventional levels, while the less likely to attract investments in soil-improving measures such as
sharecropping variable is negative, but not significant. It is significant mulch and organic manure, although fixed-rent farmers were more
to note that the investments considered are either land-conserving or likely to invest in yield increasing inputs such as mineral fertilizers.
productivity-enhancing inputs, and ownership tends to positively The positive impact of better land rights on investment remained
influence investment in these productivity-enhancing measures. The unchanged when we introduced household fixed effects into the
results also indicate positive and statistically significant effects of specification. As pointed out by Jacoby and Mansuri (2008), hold-up
access to credit and extension services. Plots farther away, as well as problems in the sense of lack of full commitment on the part of
those planted with crops such as cassava, beans and plantain also landlords could be driving the findings for fixed-rent farmers. The
indicate positive and significant effects on productivity. The results positive association of tenancy duration with investment in trees,
also show that incorporating household fixed effects does not change mulch and organic manure also suggests that making temporary
the positive and significant impact of secured land rights on farm rights longer would go a long way to enhance investments in soil-
productivity. improving measures.
We also examined the impact of tenure arrangements on farm
productivity, using an instrumental variable approach. The results
Table 5 showed a positive and significant effect of tenure security on farm
Instrumental variable estimates of determinants of productivity at plot level.
productivity, a finding that reinforces the significance of tenure
Pooled cross-section Household fixed security in encouraging higher investment in soil-improving mea-
effectsa sures. Access to credit was also found to be positively related to crop
Variable Coefficient t-value Coefficient t-value productivity, suggesting that financial constraints may be a hindrance
CONSTANT 1.172⁎⁎⁎ 3.56 0.8632⁎⁎⁎ 2.71
to investments in productivity-enhancing measures. The incorpora-
OWNER 0.4719⁎⁎ 2.39 0.2687⁎⁎ 2.26 tion of household fixed effects did not change the positive and
FIXRENT − 0.2930 1.36 − 0.1262 1.28 significant impact of secured land rights on farm productivity. The
SHARECROP − 0.0176 1.18 − 0.0527 1.02 major policy implication of these findings is that, ensuring tenure
PLTSIZE − 0.2268 1.52 − 0.2479 1.61
arrangements that confer permanent or sufficiently long temporary
PLOTFERT 0.0768⁎ 1.91 0.0612⁎ 1.97
PLANTAIN 0.1816⁎⁎ 2.26 0.0481⁎⁎ 2.24 rights to cultivators would enhance investment in both soil-
CASSAVA 0.4418⁎⁎⁎ 2.49 0.3316⁎⁎ 2.16 improving and natural resource management practices. In addition,
VBEANS 0.3261⁎⁎⁎ 2.88 0.2488⁎⁎⁎ 2.87 the results provide productivity-based arguments for enhancing
FEMALE 0.0378 0.93 0.026 1.49 farmers' access to capital.
PCREDITb 0.6879⁎⁎ 2.31
EXTEN 0.2714⁎⁎ 2.46
LIVEST 0.0251 0.89 Acknowledgements
HHSIZE 0.0187⁎ 1.87
ETHNIC 0.2516 0.96 The authors have benefited significantly from the comments and
AGE − 0.0217⁎ 1.82
suggestions of three anonymous reviewers and Mark Rosenzweig. The
TWIMEA − 0.5843⁎⁎⁎ 3.48
WORASO − 0.1457 0.96 third author acknowledges financial support of the Ministerio de
AWOROPAT − 0.2163⁎⁎ 2.13 Ciencia y Tecnología, Grant ECON 2010-17020, the Barcelona GSE and
AYEREDE − 0.1372 0.76 of the Government of Catalonia. The Netherlands Organization for
DROMA 0.0178 1.38 Scientific Research (NWO) and Amsterdam International Institute for
Adjusted R2 0.267
Development (AIID) also supported the collection of data for this
F-statistic for 6.89
overidentification [0.00] study. The usual disclaimer applies.
Number of observations 560 194

The p-values for the test statistic for the validity of the exclusion restrictions for the Appendix A
instruments are given in square brackets.
a
Household and village-level variables are dropped due to the inclusion of Demonstration of Observation 5
household fixed effects.
b
Predicted values of credit from a first-stage credit regression used in the estimation.
⁎⁎⁎ Significant at 1%. If the long-run soil capital of the owner-cultivated, Sl, is above the
⁎⁎ Significant at 5%. initial value of the soil capital S0, we refer to it as case I, and if it is
⁎ Significant at 10%. below the initial value of S0 we refer to it as case II. For case I, Fig. 1B
 A. Abdulai et al. / Journal of Development Economics 96 (2011) 66–78 77

shows that owners initially apply more organic fertilizer than than graphed in Fig. 1B because the line pOr + γ1pfXOr moves upward
sharecroppers or fixed-rent tenants, if the soil improvement effect due to the fact that fXOrS is negative. However, the impact of a change in
of organic fertilizer outweighs soil degradation. Consequently, soil fXOr is likely to be far less for pOr + γ1pfXOr than for pfXOr, since γ1 is
capital increases over time and the line pfXOr evaluated at S(t) N S0, between 0 and 1. To focus the discussion on the most likely cases we
denoted by pfXOr, with S(t) N S0, moves to the left since the cross de- do not discuss case Ia here, but it can be derived easily from Fig. 1B.
rivative of fXOrS is negative. Hence, as time advances, owners reduce the For case II, the owners' objective is to depreciate soil capital, since
O 
level of organic fertilizer from X̃ Or to X̃ Or O. This reduction of the the soil improvement effect of organic fertilizer does not outweigh soil
organic fertilizer will be somehow less than graphed in Fig. 1B, be- degradation. In this case, Fig. 1B shows that the line pfXOr moves to the
cause an increase in soil capital displaces the graph of pOr + λS(− h′ + right and over time owners apply more organic fertilizer. The optimal
O
δfXOr) b pOr downwards. The reason for this change is the decrease in application rate changes from XO Or to X Or . Fig. 1B also shows that
the shadow price of soil capital λS resulting from an increase in soil sharecroppers and fixed-rent tenants apply more organic fertilizer
capital, and in the value of fXOr due to the fact that fXOrS is negative. than XO Or, the amount applied by owner-cultivated. Hence, share-
However, the impact of a change in fXOr is likely to be far less for the croppers and fixed-rent tenant may either deplete soil capital,
graph of pOr + λS(− h′ + δfXOr) b pOr than for the graph of pfXOr, since the although to a lesser degree than owner-cultivated (case IIa), or they
change in fXOr is moderated in the former graph by the constant value may build up the soil capital (case IIb). To focus the discussion on the
of pOr, whereas there are no constant values in the latter graph. To most likely cases, a detailed discussion of cases IIa and IIb is not
focus on the main issues this movement is not presented. presented here. Following the same line of arguments these cases can
For case I, we also observe in Fig. 1B that sharecroppers and fixed- be derived easily from Fig. 1B.
rent tenants apply less organic fertilizer. Therefore, these farmers may With respect to optimal long-run application of mineral fertilizer
either build up the soil capital as well but to a lower degree than Fig. 1A shows for case I (increase of soil capital by the owner) that
O
owners (case Ia), or they deplete soil capital (case Ib). For now let us owners apply less mineral fertilizer, i.e., XO M declines to X M . If soil
consider the latter case. If sharecroppers and fixed-rent tenants capital deteriorates under sharecropping and fixed-rent tenancy (case
deplete soil capital, the function pfXOr evaluated at S(t) b S0, denoted by Ib), Fig. 1A shows that the application rate of mineral fertilizer
Sh Te
pfXOr, with S(t) b S0, moves to the right and the optimal amount of increases from XSh Te
M to X M for sharecroppers, and from XM to X M for
 Sh
organic fertilizer applied by sharecroppers increases from XSh Or to X Or fixed-rent tenants as time advances. For the sake of brevity, the
Te Te
and from XOr to X Or for the case of fixed-rent tenants. The increase in remaining cases (Ia, IIa and IIb) are not presented here but can be
organic fertilizer applied by the sharecroppers is also somehow less derived directly from Fig. 1A.

Demonstration of Observation 6
Table A1
First-stage estimations of determinants of land rights. If land is relatively scarce and fruit production is of minor
importance, the term μ2 − pFϑ will be strictly positive. In this case,
Without household With household
fixed-rent tenants tend to decrease W, from W* to WTe as it can be seen
fixed effects fixed effects
from Fig. 2. However, if fruit production plays an import role and the
Full Fixed- Share Full Fixed- Share
opportunity cost of land are small, (αμ2 − pFϑ b 0), the opposite result
rights rent cropping rights rent cropping
may arise, i.e. WTe will be situated to the right of W*. Similarly, we
CONSTANT 0.056 0.023 0.133 0.061 0.028 0.130 observe in the case of an owner-cultivated that the optimal W
(1.39) (1.10) (1.29) (1.41) (1.26) (1.18) O
decreases from W* to WO or W̃ if land is relatively scarce and fruit
PLTYEARS 0.046 − 0.116 0.072 0.049 − 0.108 0.068
(2.18) (1.61) (2.08) (2.25) (1.69) (2.17) production is of minor importance. The opposite result holds, if land is
PLOTFERT 0.212 0.247 0.077 0.203 0.236 0.106 not scarce and fruit production is quite important. However, it cannot
(2.26) (2.58) (1.23) (2.31) (2.66) (1.49) be determined whether the decrease of the owner is below or above
PLOTLOC 0.074 − 0.038 0.106 0.041 0.119 0.108
the optimal W of the tenant WTe. A similar argument applies for a
(1.93) (2.13) (1.37) (2.28) (2.24) (1.44)
PLTYEARS 0.026 0.012 0.008 0.032 0.016 0.012 sharecropper.
(1.82) (0.693) (1.47) (1.88) (1.03) (1.56)
PLTSIZE 0.220 0.286 0.083 0.218 0.247 0.079
(1.62) (1.39) (0.79) (1.73) (1.52) (1.06) References
PLTDIST 0.031 0.172 0.088 0.034 0.184 0.115
(1.85) (1.68) (2.16) (1.92) (1.76) (2.28) Angelsen, A., 1999. Agricultural expansion and deforestation: modelling the impact of
LAND INHERITED 0.121 0.128 population, market forces and property rights. Journal of Development Economics
(2.59) (2.66) 58, 185–218.
FEMALE 0.018 − 0.007 0.016 Banerjee, A., Ghatak, M., 2004. Eviction threats and investment incentives. Journal of
(0.63) (1.42) (1.17) Development Economics 74, 469–488.
ETHNIC 0.172 − 0.043 − 0.133 Banerjee, A., Gertler, P., Ghatak, M., 2002. Empowerment and efficiency: tenancy reform
(2.82) (1.98) (1.67) in West Bengal. The Journal of Political Economy 113, 151–183.
AGE − 0.017 0.012 0.018 Barbier, E.B., 2004. Explaining agricultural land expansion and deforestation in
developing countries. American Journal of Agricultural Economics 86 (5),
(1.26) (0.78) (1.07)
1347–1353.
EDUCN 0.038 0.186 0.075
Benneh, G., 1989. The dynamics of customary land tenure and agrarian systems in
(1.53) (1.05) (1.28) Ghana. In: Ghana, P. (Ed.), WCARRD; Ten Years of Follow-Up, the Dynamics of Land
IMPLTS 0.220 0.286 0.083 Tenure and Agrarian Systems in Africa, Case Studies from Ghana, Kenya,
(1.62) (1.39) (0.79) Madagascar and Togo. FAO, Rome, Italy, pp. 34–97.
HHSIZE 0.031 0.172 0.088 Besley, T., 1995. Property rights and investment incentives: theory and evidence from
(1.85) (1.68) (2.16) Ghana. The Journal of Political Economy 103, 903–937.
Village dummy Yes Yes Yes No No No Besley T. and M. Ghatak (2009). Property rights and economic development.
variables Forthcoming in D. Rodrik and M. Rosenzweig (ed.s) Handbook of Development
F-statistics 26.7 23.4 21.8 Economics V, North Holland, 2009.
[p-values] [0.00] [0.00] [0.00] Binswanger, H.P., Rosenzweig, M.K., 1986. Behavioral and material determinants of
F-test of instruments 14.72 7.86 9.13 production relations in agriculture. Journal of Development Studies 22 (3),
[p-values] [0.00] [0.00] [0.00] 503–539.
Blundell, R.W., Smith, R.J., 1989. Estimation in a class of simultaneous equation limited
Observations 560 194
dependent variable models. Review of Economic Studies 56, 37–57.
78 A. Abdulai et al. / Journal of Development Economics 96 (2011) 66–78

Brasselle, A., Frederic, G., Jean-Philippe, P., 2002. Land use security and investments Marenya, P.P., Barrett, C.B., 2007. Household-level determinants of adoption of
incentives: puzzling evidence from Burkina Faso. Journal of Development improved natural resources management practices among smallholder farmers
Economics 67, 373–418. in Western Kenya. Food Policy 32, 515–536.
Carter, M.R., Olinto, P., 2003. Getting institutions right for whom? Credit constraints and Migot-Adholla, S.E., Banneh, G., Place, F., Atsu, S., 1994. Land, security of tenure, and
the impact of property rights on the quantity and composition of investment. productivity in Ghana. In: Bruce, J.W., Migot-Adholla, S.E. (Eds.), Searching for Land
American Journal of Agricultural Economics 85, 173–186. Tenure Security in Africa. Kendall/Hunt Publishing Cy, Dubuque, pp. 97–118.
Davidson, R., Mackinnon, J.G., 1993. Estimation and Inference in Econometrics. Oxford Noronha, R., 1985. A Review of Literature on Land Tenure Systems in Sub-Saharan
University Press, New York. Africa. World Bank, Agriculture and Rural Development Department, Washington,
Deininger, K., Ali, D.A., 2008. Do overlapping land rights reduce agricultural D.C.
investment? Evidence from Uganda. American Journal of Agricultural Economics Otsuka, K., Quisumbing, A., Payongayong, E., Aidoo, J.B., 2003. Land tenure and the
90, 869–882. management of trees: the case of customary land tenure areas of Ghana.
Dorner, P., 1972. Land Reform and Economic Development. Harmondsworth; Penguin Environment and Development Economics 8, 77–104.
Books. Pinkney, T., Kimuyu, P., 1994. Land tenure reforms in East Africa: good, bad or
Ehui, S.K., Hertel, T.W., Preckel, P.V., 1990. Forest resource depletion, soil dynamics, and unimportant? Journal of African Economies 3.
agricultural productivity in the tropics. Journal of Environmental Economics and Place, F., Hazell, P., 1993. Productivity effects of indigenous land tenure systems in Sub-
Management 18 (3), 263–275. Saharan Africa. American Journal of Agricultural Economics 75, 10–19.
Feder, G., Feeny, D., 1991. Land tenure and property rights: theory and implications for Place, F., Otsuka, K., 2002. Land tenure systems and their impacts on agricultural
development policy. World Bank Economic Review 5, 135–153. investments and productivity in Uganda. Journal of Development Studies 38 (6),
Gildea, R.Y., 1964. Culture and land tenure in Ghana. Land Economics 40, 102–104. 105–128.
Goldstein, M., Udry, C., 2008. The profits of power: land rights and agricultural Quisumbing, A., Aidoo, J.B., Payongayong, E., Otsuka, K., 2001a. Agroforestry
investment in Ghana. Journal of Political Economy 116 (6), 981–1022. management in Ghana. In: Otsuka, K., Place, F. (Eds.), Land Tenure and Natural
Greene, W.H., 2008. Econometric Analysis, 6th Edition. Prentice Hall, Upper Saddle Resource Management: a Comparative Study of Agrarian Communities in Asia and
River, NJ. Africa. John Hopkins University Press, Baltimore, Maryland, pp. 53–96.
Harrison, P., 1987. The Greening of Africa. Paladin Grafton Books, London. Quisumbing, A., Payongayong, E., Aidoo, J.B., Otsuka, K., 2001b. Women's land rights in
Jacoby, H.G., Mansuri, G., 2008. Land tenancy and non-contractible investment in rural the transition to individualized ownership: implications for the management of
Pakistan. Review of Economic Studies 75, 763–788. tree resources in Western Ghana. Economic Development and Cultural Change 50,
Jacoby, H.G., Mansuri, G., 2009. Incentives, supervision, and sharecropper productivity. 157–182.
Journal of Development Economics 88, 232–241. Rivers, D., Vuong, Q.H., 1988. Limited information estimators and exogeneity tests for
Jacoby, H.G., LI, G., Rozelle, S., 2002. Hazards of expropriation: tenure insecurity and simultaneous probit models. Journal of Econometrics 39 (3), 347–366.
investment in rural China. The American Economic Review 92, 1420–1447. Wooldridge, J.M., 2002. Econometric Analysis of Cross Section and Panel Data. MIT
Lee, L., 1992. Amemiya's generalized least squares and tests of overidentification in Press, Cambridge, Massachusetts.
simultaneous equation models with qualitative or limited dependent variables.
Econometric Reviews 11, 319–328.