Fr

RSa

ob

c

U

a

ARRA

KFAPMB

I

porrm(

(co1t+Tte

r

0h

Land Use Policy 34 (2013) 168– 175

Contents lists available at SciVerse ScienceDirect

Land Use Policy

jou rn al h om epa ge: www.elsev ier .com/ locate / landusepol

armers’ decisions to adapt to climate change under various propertyights: A case study of maize farming in northern Benin (West Africa)

osaine N. Yegbemeya,b,∗, Jacob A. Yabib, Silvère D. Tovignanb, Geoffroy Gantolib,ènakpon E. Haroll Kokoyeb,c

Institute of Project and Regional Planning, Faculty of Agricultural Sciences, Nutritional Sciences, and Environmental Management, Justus-Liebig Universityf Giessen, Senckenbergstrasse 3, D-35390, GermanyDepartment of Agricultural Economics and Rural Sociology, Faculty of Agronomy, University of Parakou, BP 123 Parakou, BeninInstitute of Agriculture and Market Research, Faculty of Agricultural Sciences, Nutritional Sciences and Environmental Management, Justus-Liebigniversity of Giessen, Senckenbergstrasse 3, D-35390, Germany

r t i c l e i n f o

rticle history:eceived 4 December 2012eceived in revised form 9 January 2013ccepted 9 March 2013

eywords:

a b s t r a c t

Making the assumption that property rights might determine whether farmers adopt particular strate-gies, this study aims at modelling farmers’ decisions to adapt to climate change by focusing on theirproperty rights – declined as institutional arrangements on land and rights on land – as well as theirsocio-economic and demographic characteristics. The case study took place in northern Benin (WestAfrica). In this zone, 308 farmers producing maize and adapting to climate change were randomly sam-

armers’ decisionsdaptation to climate changeroperty rightsultivariate Probit model

enin

pled. The study was conducted by a survey method on respondents using structured interviews basedon a questionnaire. A simultaneous modelling using a Multivariate Probit (MVP) model highlighted thatsocio-economic and demographic characteristics, institutional arrangements on land, and rights on landdetermine the farmers’ decisions to adapt to climate change. The land ownership has a positive effecton the decision to adopt any adaptation strategy. Subsequently, securing farmers’ property rights wouldhelp to enhance their capacity to adapt to climate change.

ntroduction

Defined as fluctuations in the patterns of climate over longeriods (Ngaira, 2007), climate change is one of the most seri-us threats which weigh down on the sustainability of naturalesources. Moreover, agriculture which mainly relies on envi-onmental factors in developing countries will be one of theost vulnerable sectors of the global economy to climate change

Kurukulasuriya et al., 2006).In West Africa, many studies predicted either increases

Haarsmaa et al., 2005) or decreases (Held et al., 2005) in rainfall foroming years. In Benin, studies revealed that rainfall will stay morer less stable (−0.2%) in the southern part, but will reduce about3–15% in the northern part by 2100 (MEHU, 2011). At the sameime, simulations also predicted a temperature increase between2.6 ◦C and +3.2 ◦C in 2100 for the whole country (MEHU, 2011).

hese changes are expected to have differential impacts on agricul-ural productivity, food security, and agricultural livelihoods (Dixont al., 2001; Dinar et al., 2008). In Benin, climate change is already

∗ Corresponding author. Tel.: +49 179 1868152; fax: +49 6419937319.E-mail addresses: yrosaine@hotmail.fr,

osaine.n.yegbemey@agrar.uni-giessen.de (R.N. Yegbemey).

264-8377/$ – see front matter © 2013 Elsevier Ltd. All rights reserved.ttp://dx.doi.org/10.1016/j.landusepol.2013.03.001

© 2013 Elsevier Ltd. All rights reserved.

acting negatively on yields and production (Aho et al., 2006), pre-dicted to decrease about 5–20% (Paeth et al., 2008).

To sustain their livelihoods in this context of climate change,farmers have developed and implemented adaptation strategies.Indeed, throughout the literature, various adaptation strategies toclimate change are reported and the most common are: use ofnew crop varieties and livestock species that are more suited todrier conditions, irrigation, crop diversification, mixed crop live-stock farming systems, changing planting dates, diversifying fromfarm to non-farm activity, increased use of water and soil conserva-tion techniques, changed use of capital and labour and shading andsheltering/tree planting (Bradshaw et al., 2004; Maddison, 2006;Nhemachena and Hassan, 2007; Deressa et al., 2009; Till et al., 2010;Hisali et al., 2011).

Many studies explored the key determinants of farmers’ choicesin term of adaptation to climate change. But, very few papers havehighlighted the role of property rights in the process of adaptationto climate change. Running a Heckman’s sample selection modelfor understanding why farmers fail to respond to climate change,Maddison (2006) used among other explanatory variables the pro-

portions of land owned, rented out, under sharecropping, undercommunity ownership, rented, and borrowed. Even if this studystated that land tenure has little impact on the propensity of farm-ers to adapt, it has also reported that when the proportion of land

R.N. Yegbemey et al. / Land Use Policy 34 (2013) 168– 175 169

Study

btapstB

M

S

ef8Bit(avm

TS

S

D

Fig. 1.

orrowed increases, the probability of no adaptation measure beingaken by farmers also increases, raising the issue of property rightsnd adaptation to climate change. Making the assumption thatroperty rights might determine whether farmers adopt particulartrategies, this study aims at modelling the farmers’ adaptationso climate change by focusing on their property rights in northernenin (West Africa).

aterials and methods

tudy zone and database

The study took place in northern Benin. Out of the 8 agro-cological zones in Benin (cf. PANA, 2007: 17), this region coversour zones and 73% of the whole country. It is located between◦30′ and 12◦20′ North latitude, and 1◦00′ and 3◦90′ East longitude.esides, the area is considered as the basket of food and cash crops,

mplying the major role of agriculture in the livelihoods of popula-ion in the study zone. The study took into account 4 municipalities

one per agro-ecological zone): Malanville, Banikoara, Bembèrèkè,nd Natitingou (Fig. 1). In each municipality, two representativeillages were selected after preliminary interviews with key infor-ants (local leaders, heads of farmers’ organisations) and according

able 1ample structure.

Zone Municipality Village Vill

1 Malanville Koara-tédji 35Issene 18

2 Banikoara Bonhanrou 52Ounet 22

3 Bembèrèkè Guéré 25Pédarou 16

4 Natitingou Takonta 14Pam-Pam 16

Study zone – – 201

ource: Authors’ compilation.a Source: INSAE: Cahiers des villages et quartiers de ville: Départements de l’Alibori, de l’Airection des Etudes Démographiques. National Report, Cotonou, 2004.

area.

to the perception of climate change, the importance of maize pro-duction, and the accessibility to the village. The villages selectionwas done with the support of agricultural extension officers ofCeCPA (Centre Communal de Promotion Agricole).

The research units were farmers adapting to climate change.Because maize is predicted to be more affected by climate change(MEHU, 2011), we focus only on maize farms. Thus, after a rapidcensus survey of all maize producers in each selected village, 308farmers producing maize and adapting to climate change were ran-domly sampled using the table of random numbers. Table 1 showsthe sample structure.The data collected were about the farmers’socio-economic and demographic characteristics (gender, level ofschooling, etc.), the adaptation strategies to climate change, and theproperty rights. The study was conducted by survey methods onrespondents using structured interviews based on a questionnaire.Data were analysed with STATA 11.

Modelling farmers’ adaptations to climate change

There is a growing interest in the literature on climate changeadaptation for understanding the reasons underlying the farmers’responses (adaptation strategies). For farmers, adaptation to cli-mate change requires to make choices among a set of adaptation

age’s householdsa Sample size Sampling rate (%)

9 43 11.982 38 20.88

3 39 7.462 39 17.57

2 37 14.685 40 24.24

1 30 21.287 42 25.15

1 308 15.32

tacora et de la Donga. Institut National de la Statistique et de l’Analyse Economique.

1 d Use P

oyt

mLunsTai(

a

wtoi

Im

rlarpa

a

wactpc(pdasspfoosamttlta⎧⎪⎪⎪⎪⎪⎨⎪⎪⎪⎪⎪⎩

70 R.N. Yegbemey et al. / Lan

ptions (practices and technologies) available. Therefore, the anal-sis of farmers’ adaptations to climate change aims at finding outhe important determinants of adoption decisions.

For analysing adoption decision in general and adoption of cli-ate change adaptation strategies in particular, the Multinomial

ogit (MNL) or the Multinomial Probit (MNP) models are widelysed. Both MNL and MNP are also appropriate for evaluating alter-ative combinations of adaptation strategies, including individualtrategies (Hausman and Wise, 1978; Wu and Babcock, 1998).herewith, the dependent variable is a set of strategies defineds one variable (a) with multiple modalities. Most of the modelsnclude j farmers’ socio-demographic and economic characteristicsz), giving:

i = ˛0 +∑

j˛izij + ui (1)

ith: ˛0 the constant term, ˛j the parameters to be estimated and uhe error terms. To this basic form and according to the specificitiesf each study, different independent or exogenous variables may bencluded in the model.

ntegrating property rights in climate change adaptationodelling

We considered two main sets of variables to assess propertyights: the institutional arrangements on land (w) and the rights onand (r). The institutional arrangements on land refer to the waysnd arrangements for having access to land while the rights on landefer to the rights that farmers have to use land for different pur-oses. Giving this, the easiest way for interrelating property rightsnd climate change adaptation model (Eq. (1)) is:

i = ˛0 +∑

j˛jzij +

∑m

ˇmwim +∑

m′ım′ rim′ + ui (2)

here m and m′ are the numbers of considered institutionalrrangements and rights on land respectively; ˛j, ˇm and ım′ areoefficients to be estimated. The aim of this paper is to isolatehe effect of property rights on each adaptation strategy. For thisurpose, we need a framework representing different (possiblyorrelated) separate decisions. However, the Multinomial LogitMNL) or the Multinomial Probit (MNP) frameworks represent theossible outcomes – adaptation options – viewed as one jointecision. Therefore, even though the MNL and MNP present thedvantage to deal with combinations among adaptation strategies,imple regression models for each adaptation option consideredeparately, using a Logit or Probit specification appear to be appro-riated. But, assuming that the different adaptation options areull part of farmers’ response to climate change, the adoption ofne given option or strategy might be correlated with the adoptionf another option. With this statement, a simultaneous modellingystem – seemingly unrelated regression models for instance –ppears to be more appropriated. Moreover, in such model the esti-ated coefficients will account for unobserved correlation among

he outcomes and will be therefore less biased and more efficienthan those produced by the independent-separated-models. In thisine, the dependent variables to be analysed are the farmers’ adap-ations, each decision considered as binary variable. Then for kdaptation decisions, the Eq. (2) becomes:

a1i + ˛1 +∑

j˛1jzij +∑

mˇ1mwim +∑

m′ı1m′rim′ + u1i

a2i = ˛2 +∑

j˛2jzij +∑

mˇ2mwim +∑

m′ ı2m′ rim′ + u2i

...

aki = ˛k +∑

j˛kjzij +∑

mˇkmwim +∑

m′ ıkm′ rim′ + uki

(3)

olicy 34 (2013) 168– 175

For each adaptation decision,

aki ={{

1 if ˛k + ∑j˛kjzij + ∑

mˇkmwim + ∑m′ıkm′ rim′ > 0

0 otherwise

and uk → N(

0,∑)

In this equation: ˛k are the constant terms; ˛kj, ˇkm, and ıkm′are the coefficients of zj, wm, and rm, respectively; and uk the errorterms. Let’s call A a kk × 1 vector of adaptation strategies, Z a kj × 1vector of farmers’ socio-economic and demographic characteris-tics, W a km × 1 vector of institutional arrangements on land, and Ra km′ × 1 vector of rights on land. Then, the system Eq. (3) becomes:

A = C + ZX˛ + WXˇ + RXı + U (4)

The coefficients ˛, ˇ, and ı were estimated using a Multivari-ate Probit (MVP) model. This method estimates M-equation Probitmodels by using the Simulated Maximum Likelihood (SML) method(Cappellari and Jenkins, 2003). From the estimated coefficients andtheir levels of significance, the factors affecting farmers’ adaptationdecisions were indentified.

Choice of the exogenous variables and hypotheses to be tested

The prospective explanatory variables considered are listed inTable 2. The selection of variables was grounded in the literatureand our field observations. Thus, the variables were selected follow-ing assumptions. The assumptions were set out regarding the broaddecision to adapt to climate change since it was too hazardous topredict the impact of one given variable on a specific adaptationstrategy.

Socio-economic and demographic characteristics (Z)- Gender: In northern Benin, agriculture is mainly performed by

men (Paraïso et al., 2011, 2012) who hold the largest farms.Women usually have their own small plots of land where theyproduce vegetables for household consumption. Subsequently,denoting women as 0 and men as 1, we expect a positive relation-ship between gender and farmers’ decision to adapt to climatechange.

- Level of education: Educated farmers are more likely to respondto climate change by making at least one adaptation strategy(Maddison, 2006). Following this statement, the more farmersare educated, the more they perceive climate change and developadaptation strategies. Therefore, we expect that the level of edu-cation will have a positive effect on farmers’ decision to adapt toclimate change.

- Experience in agriculture:According to Ofuoku (2011), farmerswho have many years of farming experience have interactedmuch more with the climate in relation to their farming activities,and therefore, have good knowledge of environmental factors asthey relate to their farming operations. Thus, the more farmershave experience in agriculture, the more they are more likely toadapt. Thus we hypothesise that the experience in agriculturewill be positively correlated with the farmers’ decision to adaptto climate change.

- Household’s size: We assume that bigger households have morelabour available for performing agricultural activities. So wehypothesise that bigger households have more possibilities toadapt to climate change in term of labour availability. This ismainly relevant for adaptation strategies (double sowing instead

of single sowing for instance) requiring more labour input.

- Contact with extension service: According to Maddison (2006),farmers who have received extension advice are likely to adaptto climate change. Thus, we expect that contact with extension

R.N. Yegbemey et al. / Land Use Policy 34 (2013) 168– 175 171

Table 2Explanatory variables considered in the model.

Variables Typesa Modalities Expected signs

Socio-economic and demographic characteristics (Z)Gender D 0 = Female; 1 = Male +Education level (years) C – +Experience in agriculture (years) C – +Household’s size (person) C – +Contact with extension service D No = 0; Yes = 1 +Access to credit D No = 0; Yes = 1 +

Institutional arrangements on land (W)Inheritance D No = 0; Yes = 1 +Gifting D No = 0; Yes = 1 −Renting D No = 0; Yes = 1 +Purchasing D No = 0; Yes = 1 +

Rights on land (R)Ownership D No = 0; Yes = 1 +Right of cutting trees D No = 0; Yes = 1 −Right of selling D No = 0; Yes = 1 −Right of leasing/renting D No = 0; Yes = 1 −

Right of gifting D No = 0; Yes = 1 −Right of giving as inheritance D No = 0; Yes = 1 +

S

-

I-

-

-

-

R

-

ource: Authors’ specifications.a Types: D = discontinuous variables; C = continuous variables.

services especially the ones related to climate change adaptationwill affect positively the farmers’ decision to adapt to climatechange.

Access to credit: Access to credit is well known as a key determi-nant of farmers’ decisions (Shahidur et al., 2002). We hypothesise,that the access to credit enhances farmers’ financial capital, mak-ing them able to adapt to climate change. This is particularlyrelevant for adaptation strategies (shortening of the productionperiod which implies an additional need of labour (by hiringfor instance) for performing activities on time) requiring moreinputs (labour and fertilisers for instance).

nstitutional arrangements on land (W) Inheritance: Farmers protect their patrimony (land for instance)and keep it productive as much as possible. As well, they are will-ing to conserve land for future generation. Thus, we assume thatinheritance will have positive impact on the decision to adapt toclimate change, particularly regarding land conservation issue.

Gifting: Most of the time, land gifting happens when a farmerwants to thank his fellow or someone else. In this context, landgifting does not aim at satisfying a need of land for produc-tion. Subsequently, accessing land through gifting arrangementis expected to be negatively correlated with the decision to adaptto climate change.

Renting: When land is accessed through renting arrangement, itmeans that the users have to give it back to the owners sooner orlater. As well, the users have to keep land productive; otherwise,the owners may take it away from them. In the particular case ofshare cropping arrangement, the users have to develop strategiesaiming at using all opportunities (from both social-economic andenvironmental point of views) for maximising output. Therefore,while accessing land by renting, farmers are likely to adapt toclimate change.

Purchasing: Purchasing land is a form of investment. When land isacquired through purchase, farmers would want to optimise theinvestment. However, most of the time, farmers sell useless land.When the purchaser realise that, he may be disappointed and lesswilling to invest, implying to adapt to climate change.

ights on land (R)

Land ownership: Land ownership broadly entitles a full righton land, implying a certain level of land security. According to

Quan (2006), the more land access is secured, the more farm-ers invest. Thus, we hypothesise that land ownership will bepositively correlated with farmers’ decision to adapt to climatechange.

- Right of cutting trees: In northern Benin, only the owner of landhas the right of cutting trees in the field. However, there is a con-trast between the need to adapt to climate change and the fact ofcutting threes (threes can be used as adaptation to strong windsand agro-forestry for instance). Therefore, we expect a negativerelationship between these variables.

- Right of selling: Farmers are willing to sell land that is no more pro-ductive for them. Moreover, farmers often state that one shouldnot sell land (fields for instance). Subsequently, we assume thatfarmers are less willing to adapt to climate change regarding theright of selling land.

- Right of leasing/renting: According to Maddison (2006), whenthe proportion of land borrowed increase, the probability of noadaptation measure being taken by farmers also increase. Giv-ing this, we assume that the right of leasing/renting out landwill negatively impact on farmers’ decision to adapt to climatechange.

- Right of gifting: Farmers usually make gifting of land that is becom-ing useless for them. This is, we must state it, a rational behaviourfrom an economic point of view. Thus, having the right to giftthe land could be negatively correlated with farmers’ decision toadapt to climate change.

- Right of giving land as inheritance: As previously in the set ofinstitutional arrangements on land, we consider that farmers arewilling to conserve land for future generation. Thus, we expecta positive impact of the right of giving land as inheritance onfarmers’ decision to adapt to climate change.

Results

Farmers’ socio-economic and demographic characteristics

The descriptive statistics of farmers’ socio-economic and demo-graphic characteristics (Table 3) show that respondents are mainly

men aged about 40 years with 22 years of experience in agricul-ture on average. The respondents are all head of their household.The average household size is 13 persons. The level of education islow (5 years of schooling in average) as well as the farmers’ access

172 R.N. Yegbemey et al. / Land Use Policy 34 (2013) 168– 175

Table 3Socio-economic and demographic characteristics (N = 308).

Qualitative variables Frequency Percentage

Gender – –Female 21 06.82Male 287 93.18Level of education 150 48.70Contact with extension service 177 57.47Access to credit 66 21.43

Quantitative variables Mean Standard deviation

Age (in years) 39.52 11.97Level of education (in year) 4.68 4.06Experience in agriculture (in years) 21.77 11.65

S

te

F

mcaff

fi

TA

NS

Table 5Institutional arrangements on land (N = 308).

Arrangements Frequency Percentage

Inheritance 229 74.35Gift 41 13.31Renting 11 3.57Purchasing 08 2.60

Source: Authors’ calculations.

Table 6Farmers’ rights on land (N = 308).

Rights on land Frequency Percentage

Ownership 267 86.69Right of cutting trees 266 86.36Right of selling 197 63.96Right of leasing/renting 227 73.70Right of gifting 212 68.83

one covariance of the error term is statistically significant, implyingthat the equations in the model are connected. Subsequently, the

Household size 12.59 08.45

ource: Authors’ calculations

o credit (21%). Finally, about half of farmers are in contact withxtension service.

armers’ adaptations to climate change

Many strategies are developed by farmers for adapting to cli-ate change. We found out four groups of adaptation strategies

urrently used by farmers producing maize in northern Benin. It isbout the crops diversification strategies, the farming practices orarming calendar adjustment, the land use strategies, and the otherorms of strategies.

Option 1. Adaptation to climate change through crops diversificationstrategies: In this group we find strategies such as the use of dif-ferent crops, different varieties, including varieties with shortercycle.Option 2. Adaptation to climate change through farming practicesand farming calendar adjustments: The farming practices developedas adaptation to climate change are the crops rotation or associa-tion, the micro-level irrigation, the double sowing, the changes inthe doses of inputs (fertilisers and herbicides for instance), and thesoil conservation strategies. The farming calendar adjustments areabout changes in dates of cropping activities.Option 3. Adaptation to climate change through land use: Thechanges in the land allocation among different crops and the trans-fer from one production site to another in the surrounding areasare strategies grouped under this option.Option 4. Adaptation to climate change through other strategies: Inthis group, we find all strategies which are not mentioned above.For instance, change of activity (agriculture to non-farm activity),migration to another agro-ecological zone, prayers, and access tocredit are other strategies used by farmers to adapt to climatechange.

Among these strategies, the crops diversification strategies, thearming practices and farming calendar adjustment are the mostmplemented by both men and women (Table 4).

able 4bsolute and relative frequencies of adaptation options (N = 308).

Adaptation Women Men Total

Crops diversification 18 (5.8) 250 (81.2) 268 (87.0)Farming practices and farming

calendar19 (5.8) 265 (86.0) 284 (92.2)

Land use 5 (1.6) 40 (13.0) 45 (14.6)Other adaptations 19 (6.2) 200 (64.9) 219 (71.1)

ote: Values in brackets are relative frequencies.ource: Authors’ calculations.

Right of giving as inheritance 260 84.42

Source: Authors’ calculations.

Farmers’ property rights

Institutional arrangements on landFarmers have access to land through four main institutional

arrangements: inheritance, gift, renting, or purchasing (Table 5).Inheritance is the main arrangement governing land access in thestudy zone. The land purchase for agricultural purpose is very rare.Indeed, most of the time, land is purchased by foreigners-peoplefrom other municipalities-for buildings. The classical land rentingis also rare because farmers do not rent land among them, only forforeigners willing to produce in the area. Therefore, the traditionalrenting system among farmers is without share of output or simplylike a gift.

Rights on landThe access to land entitled farmers different rights. This study

focused on the farmers’ ownership as well as their rights for cut-ting trees in the field, selling land, leasing or renting land, giftingland, and giving land as inheritance. From Table 6, it comes out thatfarmers have mainly ownership right and right for cutting trees andgiving land as inheritance.

Property rights and climate change adaptation

The results of the Multivariate Probit (MVP) model–Wald chi2(64) = 98.95; log likelihood = −268.59; Probability > chi2 = 0.0033 –indicate the relevance of the model for explaining the role ofproperty rights in farmers’ decisions to adapt to climate change.Moreover, the Likelihood ratio test (Table 7) indicates that at least

estimated coefficients account for unobserved correlation amongthe adaptation decisions. Coming back to the Mulvivariate Probit

Table 7Covariance of the error terms and likelihood ratio test.

rho Coefficient Standard error P > |z|rho21 0.55 0.17 0.002rho31 0.08 0.22 0.705rho41 0.53 0.16 0.001rho32 0.61 0.18 0.001rho42 0.45 0.14 0.002rho43 0.21 0.15 0.149

Likelihood ratio test of: rho21 = rho31 = rho41 = rho32 = rho42 = rho43 = 0.Chi2 (6) = 26.50, probability > chi2 = 0.0002.Source: Authors’ calculations.

R.N. Yegbemey et al. / Land Use Policy 34 (2013) 168– 175 173

Table 8Results of the Multivariate Probit (MVP) model.

Variables Crops diversification Farming practices andfarming calendaradjustments

Land use strategies Other strategies

Coefficient P > z Coefficient P > z Coefficient P > z Coefficient P > z

Socio-economic and demographic characteristics (Z)Gender 0.947* (0.48) 0.052 0.065 (0.55) 0.906 −0.259 (0.489) 0.597 −0.808* (0.48) 0.098Level of education −0.009 (0.035) 0.784 0.002 (0.03) 0.952 −0.020 (0.03) 0.557 −0.008 (0.02) 0.756Experience 0.013 (0.01) 0.281 −0.002 (0.01) 0.825 0.030** (0.01) 0.021 0.007 (0.01) 0.400Household size 0.019 (0.01) 0.303 0.031 (0.02) 0.181 −0.014 (0.01) 0.393 0.017 (0.01) 0.204Contact extension 0.442 (0.28) 0.124 0.350 (0.31) 0.262 −0.403 (0.26) 0.135 0.944*** (0.23) 0.000Access credit 0.375 (0.37) 0.316 0.835** (0.32) 0.011 −0.234 (0.30) 0.441 −0.349 (0.24) 0.152

Institutional arrangements on land (W)Inheritance −1.015 (0.63) 0.109 0.775 (0.48) 0.109 2.073*** (0.68) 0.003 0.020 (0.41) 0.960Gifting 0.084 (0.56) 0.881 −0.072 (0.50) 0.886 1.169** (0.51) 0.022 0.261 (0.42) 0.530Renting 3.864 (2.34) 0.987 −0.477 (0.73) 0.518 1.355* (0.80) 0.093 −1.651** (0.79) 0.038Purchasing −2.531*** (0.75) 0.001 4.439 (1.96) 0.982 1.834*** (0.67) 0.007 −1.484** (0.75) 0.048

Rights on land (R)Ownership 0.613 (0.72) 0.400 0.055 (0.59) 0.926 0.631 (0.60) 0.297 1.262** (0.62) 0.043Cutting trees −0.708 (0.72) 0.328 −1.094* (0.63) 0.084 −0.273 (0.45) 0.548 −0.313 (0.44) 0.483Selling −0.799* (0.43) 0.068 −0.502 (0.43) 0.253 −0.073 (0.34) 0.83 0.370 (0.30) 0.231Leasing/renting −0.421 (0.62) 0.499 −0.422 (0.58) 0.467 −0.464 (0.43) 0.289 −0.403 (0.36) 0.276Gifting 0.318 (0.37) 0.393 0.583 (0.36) 0.112 −0.145 (0.32) 0.658 −0.033 (0.29) 0.908Inheritance 1.387** (0.65) 0.033 0.090 (0.60) 0.881 0.991* (0.53) 0.062 −0.033 (0.42) 0.939Constant −0.011 (0.71) 0.987 1.615* (0.89) 0.070 −1.241* (0.71) 0.083 2.098** (0.84) 0.014

Wald chi2 (64) = 98.95; log likelihood = −268.59897***; probability > chi2 = 0.0033.NB: the values in bracket are the standard-errors.Source: Authors’ estimations.

* Significant at 10% (p < 0.10).

mtd

-

-

-

** Significant at 5% (p < 0.05).*** Significant at 1% (p < 0.01).

odel results, socio-economic and demographic characteristics,he institutional arrangements on land, and property rightsetermine the farmers’ adaptation to climate change (Table 8).

Socio-economic and demographic characteristics adaptation to cli-mate change: Gender determines farmers’ decision to adopt cropsdiversification strategies or other adaptation strategies. Menare predominant in the sample (93%) and mainly adopt cropsdiversification strategies, farming practices and farming calen-dar adjustments and other adaptations. As well, adaptations suchas crops diversification and farming practices adjustment requiremore investment in capital and land, and men traditionally havemore access to land and capital. The farmer’s experience in agri-culture determines the adoption of land use adaptation strategies.Indeed, deciding on how to allocate land among different cropsor change the land use system itself requires a strong knowl-edge that farmers can but learn through experience. Contact withextension service determines the adoption of other strategies(access to credit for instance). Indeed, besides the technical sup-port, extension services stimulate farmers to access credit evenif the rate of access to credit is still low. The expected resultis to enhance farmers’ capacities to adapt to climate change. Inthis way, the access to credit determines the adoption of farmingpractices and farming calendar adjustments. Finally, the school-ing level and the household size do not determine the famers’adoption decisions.

Institutional arrangements on land and adaptation to climatechange: All institutional arrangements on land – inheritance,gifting, renting, purchasing – influence the land use strategiesadoption. In addition, renting and purchasing negatively deter-mine the adoption of other strategies in one hand and the

purchasing the adoption of crop diversification strategies in theother hand.Property rights and adaptation to climate change: The land owner-ship has a positive effect on any decision to adapt to climate

change although the effect is only significant in the case of otheradaptation strategies. The right of cutting or planting trees in thefield has a negative effect on any adoption decision of adaptationstrategies. But this effect is only significant in the case of adop-tion of farming practices and farming calendar adjustments. Inthe study area, only the owners of land have right to cut or planttrees. Then, as the ownership, the right of cutting trees is expectedto have positive effect on farmers’ decision of adaptation. That is,the negative effects of cutting trees right seem to be in contra-diction with the ownership effect. This result can be explained bythe fact that farmers, even if they are owners of land, do not wantto cut trees in the field for many reasons such as for securing theirland, avoiding soil erosion, etc. To some extent, the right no to cuttrees stands for the one which encourage adaptation to climatechange. The rights of leasing or renting out land and gifting landhave no significant effect on the adoption of adaptation strategies.The right of giving land as inheritance has a positive and signifi-cant effect on the adoption of crop diversification strategies andland use strategies. This right assumes that the farmer is the onlyone owner of land.

Discussion

Adaptation is one of the key policy options for reducing the neg-ative impact of climate change (Adger et al., 2003; Kurukulasuriyaand Mendelsohn, 2008). Following Mendelsohn et al. (1994)farmers will be especially hard hit if they do not adjust at all to newclimates by implementing adaptation strategies. These strategiesrefer to adjustments in natural or human systems in response toactual or expected climatic stimuli (IPCC, 2001). They help farmersto achieve their food, income and livelihoods security (Kandlinkar

and Risbey, 2000). The adaptation strategies reported in the liter-ature are manifold. They include crops diversification strategies(cf, Maddison, 2006; Nhemachena and Hassan, 2007; Deressaet al., 2009), farming practices or farming calendar adjustments,

1 d Use P

tGN

tsmFenisihm(tasba

rdlatao

drpprYusllcfapataiicafaac

C

cadaee

74 R.N. Yegbemey et al. / Lan

he land use strategies (cf. Maddison, 2006; Deressa et al., 2009;nanglè et al., 2012), and the other strategies such as prayers (cf.hemachena and Hassan, 2007).

The farmers’ decision to implement one or many of these adap-ation strategies is influenced by various exogenous factors. Theocio-economic and demographic characteristics are one of theost important factors determining farmers’ agricultural choices.

or instance, from their experience, farmers learnt to know theirnvironment and even anticipate on events such as rains, begin-ing of seasons, etc. Thus, the higher the experience in agriculture

s, the more farmers are able to set out climate change adaptationtrategies. This supports the statement of Ofuoku (2011) accord-ng to whom farmers who have many years of farming experienceave good knowledge of climatic factors. Subsequently, they areore likely to adapt to climate change since, following Awoyé et al.

2012), farmers’ perception of climate change shapes their adapta-ion strategies. In addition, focusing on climate change uncertainty,gricultural extension appears to be a way to support farmers’ deci-ion in the process of climate change adaptation. Indeed, as statedy Maddison (2006), farmers who have enjoyed extension advicere likely to adapt to climate change.

According to Cotula et al. (2004), the access to land and the landights will determine to many extends the direction of agriculturalevelopment. In the same line, the institutional arrangements on

and and the farmers’ rights on land influence farmers’ decision todapt to climate change. Even though Maddison (2006) reportedhat land tenure has little impact on the propensity of farmers todapt, farmers set out the land management strategies dependingn the way of access to the aforesaid land.

The land ownership is an important factor in the farmer’secision to adapt to climate change. Economic analysis has longecognised the importance of secure property rights for growth andoverty reduction (Quan, 2006). Moreover, the literature provideslenty of evidence that farmers are not likely to be neutral toisk and actually tend to be risk averse agents (Serra et al., 2006;esuf and Bluffstone, 2007). Putting this in the context of climaticncertainty, farmers are not willing to implement adaptationtrategies – that may require more inputs investment – when theirand is not secure or they do not have full rights on the aforesaidand. According to Kokoye et al. (2013), secure property rightsould be used as an incentive in investing in agriculture. Likewise,ollowing Ghei (2008), when property rights are not secure, there is

non-zero and positive probability of expropriation. This positiverobability of expropriation acts as a random tax on the property,nd reduces the expected return on any investment that enhanceshe value of the property, and acts as a disincentive to invest. As

result, secure access to land creates incentives for the user tonvest labour and other resources in it so as to maintain or enhancets value and sustain its productivity (Quan, 2006), even in theontext of climate change. Finally, social considerations accountlso in farmers’ decision. Indeed, since a farmer can make hisamily inheritance of land, he is more willing to care about it bydopting practices that could help keeping the land productive in

context of climate change. Likewise, when farmer knows that heould sell land, he is less willing to adopt certain strategies.

onclusion

Our study showed that both socio-economic and demographicharacteristics and property rights – declined in institutionalrrangements on land and rights on land – influence the farmers’

ecisions to adapt to climate change. Talking about climate changedaptation and property rights, the land ownership has a positiveffect on any decision of adaptation. It acts as an incentive for farm-rs investments in term of land security. Subsequently, securing

olicy 34 (2013) 168– 175

farmers’ property rights would help to enhance their capacity toadapt to climate change.

Acknowledgements

The authors are grateful to the German Academic ExchangeService (DAAD) for financing this research. Thanks to GhislainAihouton and Michel Atchikpa for helping in data collection, and tothe anonymous reviewers for their relevant suggestions and com-ments. Our acknowledgements also go to our scientific advisor Prof.Dr. Siegfried Bauer for his support.

References

Adger, W.N., Huq, S., Brown, K., Conway, D., Hulme, M., 2003. Adaptation to climatechange in the developing world. Progress in Development Studies 3, 179–195.

Aho, N., Boko, M., Afouda, A., 2006. Evaluation concertée de la vulnérabilité auxvariations actuelles du climat et aux phénomènes météorologiques extrêmes.PANA/Bénin, 93.

Awoye, O.H.R., Agbossou, E.K., Yegbemey, R.N., Kokoye, S.E.H., 2012. Scientific andLocal Knowledge on Climate Change in West Tropical Africa: Do Farmers’ Per-ceptions Fit the Measured Changes? Resilience of Agricultural Systems AgainstCrises, Book of Abstract. Proceedings Tropentag 2012. p. 423. ISBN 978-3-95404-215-9.

Bradshaw, B., Dolan, H., Smit, B., 2004. Farm-level adaptation to climatic variabilityand change: crop diversification in the Canadian prairies. Climatic Change 67,119–141.

Cappellari, L., Jenkins, S.P., 2003. Multivariate probit regression using simulatedmaximum likelihood. The Stata Journal 3 (3), 278–294.

Cotula, L., Toulmin, C., Hesse, C., 2004. Land Tenure and Administration in Africa:Lessons of Experience and Emerging Issues. IIED, London.

Deressa, T.T., Hassan, R.M., Ringler, C., Alemu, T., Yesuf, M., 2009. Determinants offarmers’ choice of adaptation methods to climate change in the Nile Basin ofEthiopia. Global Environmental Change 19 (2009), 248–255.

Dinar, A., Hassan, R., Mendelson, R., Benhin, J., 2008. Climate Change and Agriculturein Africa: Impact Assessment and Adaptation Strategies. Earthscan/Centre forEnvironmental Economics and Policy in Africa (CEEPA), London, p. 189, ISBN 13:978 1 84407 547 8.

Dixon, J., Gulliver, A., Gibbon, D., 2001. Farming Systems and Poverty:Improving Farmers’ Livelihoods in a Changing World. FAO and WorldBank, Rome and Washington, DC, Available online: ftp://ftp.fao.org/docrep/fao/004/ac349e/ac349e00.pdf

Ghei, N., 2008. Institutional Arrangements, Property Rights and the Endogen-ity of Comparative Advantage. ExpressO, Available at: http://works.bepress.com/nita ghei/1

Gnanglè, P.C., Yabi, J.A., Yegbemey, N.R., Glèlè Kakaï, L.R., Sokpon, N., 2012. Rentabil-ité économique des systèmes de production des parcs à Karité dans le contextede l’adaptation au changement climatique du Nord-Bénin. African Crop ScienceJournal 20 (s2), 589–602.

Haarsmaa, R.J., Selten, F.M., Weber, S.L., Kliphuis, M., 2005. Sahel rainfall variabilityand response to greenhouse warming. Geophysical Research Letters 32, 1–4.

Hausman, J., Wise, D., 1978. A conditional probit model for qualitative choice: dis-crete decisions recognizing interdependence and heterogeneous preferences.Econometrica 46, 403–426.

Held, I.M., Delworth, T.L., Lu, J., Findell, K.L., Knutson, T.R., 2005. Inaugural article:simulation of Sahel drought in the 20th and 21st centuries. Proceedings of theNational Academy of Sciences 102, 17891–17896.

Hisali, E., Birungi, P., Buyinza, F., 2011. Adaptation to climate change in Uganda:evidence from micro level data. Global Environmental Change 21, 1245–1261.

IPCC, 2001. Climate Change 2001. Impacts, Adaptation and Vulnerability. Con-tribution of Working Group II to the Third Assessment Report of theIntergovernmental Panel on Climate Change. WMO for Intergovernmental Panelon Climate Change. Cambridge University Press, Cambridge, UK.

Kandlinkar, M., Risbey, J., 2000. Agricultural impacts of climate change: if adaptationis the answer, what is the question? Climatic Change 45, 529–539.

Kokoye, S.E.H., Tovignan, D.S., Yabi, J.A., Yegbemey, R.N., 2013. Economet-ric modeling of farm household land allocation in the municipalityof banikoara in northern benin. Land use policy 34 (2013), 72–79,http://dx.doi.org/10.1016/j.landusepol.2013.02.004.

Kurukulasuriya, P., Mendelsohn, R., 2008. A Ricardian analysis of the impact ofclimate change on African cropland. African Journal of Agricultural and ResourceEconomics 2 (1), 1–23.

Kurukulasuriya, P., Mendelsohn, R., Hassan, R., Benhin, J., Diop, M., Eid, H.M.,Fosu, K.Y., Gbetibouo, G., Jain, S., Mahamadou, A., El-Marsafawy, S., Ouda, S.,Ouedraogo, M., Sène, I., Maddision, D., Seo, N., Dinar, A., 2006. Will African agri-culture survive climate change? World Bank Economic Review 20 (3), 367–388.

Maddison, D., 2006. The Perception and Adaptation to Climate Change in Africa.CEEPA Discussion Paper No. 10. Centre for Environmental Economics and Policyin Africa, University of Pretoria, South Africa.

MEHU, 2011. Deuxième Communication Nationale de la République du Bénin surles Changements Climatique. MEHU, Benin, Cotonou, p. 168.

d Use P

M

N

N

O

P

P

P

R.N. Yegbemey et al. / Lan

endelsohn, R., Nordhaus, W., Shaw, D., 1994. The impact of global warmingon agriculture: a Ricardian analysis. American Economic Review 84, 753–771.

gaira, J.K.W., 2007. Impact of climate change on agriculture in Africa by 2030.Scientific Research and Essays 2 (7), 238–243.

hemachena, C., Hassan, R., 2007. Micro-Level Analysis of Farmers’ Adaptation toClimate Change in Southern Africa. IFPRI Discussion Paper No. 00714. Interna-tional Food Policy Research Institute, Washington, DC.

fuoku, A.U., 2011. Rural farmers’ perception of climate change in central agricul-tural zone of Delta state, Nigeria. Indonesian Journal of Agricultural Science 12(2), 63–69.

aeth, H., Capo-Chichi, A., Endlicher, W., 2008. Climate change and food security intropical West Africa – a dynamic–statistical modeling approach. Erdkunde 62(2), 101–115.

araïso, A., Yabi, A.J., Sossou, A., Zoumarou-Wallis, N., Yegbemey, R.N., 2012.Rentabilité économique et financière de la production cotonnière à ouaké

au nord–ouest du Bénin. Annales des Sciences Agronomiques 16 (1),91–105.

araïso., A., Sossou, A.C.G., Yegbemey, R.N., Biaou, G., 2011. Analyse de la rentabilitede la production du fonio (Digitaria Exilis S.) Dans la commune de boukombé aubenin. Rech. Sci. Univ. Lomé (Togo) 13 (1), 27–37, Série A.

olicy 34 (2013) 168– 175 175

Quan, J., 2006. Land access in the 21st century: issues, trends, linkages andpolicy options. Food and Agriculture Organization of the United Nations, Liveli-hood Support Programme (LSP). Access to Natural Resources Sub-Programme.LSP Working Paper 24, p. 87. Available online http://www.fao.org/es/esw/lsp/cd/img/docs/LSPWP24.pdf

Serra, T., Zilberman, D., Goodwin, B.K., Featherstone, A.M., 2006. Effects of decou-pling on the mean and variability of output. European Review of AgriculturalEconomics 33, 269–288.

Shahidur, R., Sharma, M., Zeller, M., 2002. Micro-lending for small farmers inBangladesh: does it affect farm households’ land allocation decision? Marketsand Structural Studies Division. International Food Policy Research Institute.MSSD discussion paper No. 45.

Till, B., Artner, A., Siebert, R., Sieber S., 2010. Micro-level Practices to Adapt to ClimateChange for African Small-scale Farmers. A Review of Selected Literature. Envi-ronment and Production Technology Division. IFPRI Discussion Paper 00953, p.28.

Wu, J., Babcock, B.A., 1998. The choice of tillage, rotation, and soil testing practices:economic and environmental implications. American Journal of AgriculturalEconomics 80, 494–511.

Yesuf, M., Bluffstone, R., 2007. Risk aversion in low income countries, experimentalevidence from Ethiopia. IFPRI Discussion Paper 00715.