Internatonal Journal of Development and Sustanablty ISSN: 2186-8662 www.sdsnet.com/jds Volume 3 Number 9 (2014): Pages 1862-1877 ISDS Artcle ID: IJDS14032304 Reverse tenancy and dstress contractng among rural farmng households n Malaw Boja Tesfaye 1*, Abd-Khall Edrss 2 1 Lecturer, Jmma Unversty, Ethopa 2 Professor of Appled Economcs and Statstcs, Llongwe Unversty of Agrculture and Natural Resources (Bunda, LUANAR), Malaw Abstract Ths study analyzes factors assocated wth reverse tenancy and dstress contractng n the land rental market for rural households n Malaw. Land rental markets transfer land from the land rch but resource poor to the land poor but wealthy households. It s also a remedy to the problem of land scarcty due to the soarng populaton growth and the resultant hgh pressure on farm land. We employed the bnary probt model for partcpaton, and the censored (Tobt) model for the degree of partcpaton for both tenants and landlords to pn down soco-economc factors affectng reverse tenancy and dstress contractng n the land rental market usng 450 randomly sampled households across sx dstrcts. The study revealed that fxed rental contracts domnated whle sharecroppng was rare. Reverse tenancy contracts where tenants are rcher than landlords n non-land assets were common. Own land and non-land resources (famly labor, assets, tropcal lvestock unts and others), and household poor health condton were found to be sgnfcantly assocated wth land market partcpaton and the degree of partcpaton. Keywords: Land market; Resource varaton; Fxed rental; Share croppng; Fxed rental; reverse tenancy; Malaw Publshed by ISDS LLC, Japan Copyrght 2014 by the Author(s) Ths s an open access artcle dstrbuted under the Creatve Commons Attrbuton Lcense, whch permts unrestrcted use, dstrbuton, and reproducton n any medum, provded the orgnal work s properly cted. Cte ths artcle as: Tesfaye, B. and Edrss, A.K. (2014), Reverse Tenancy and Dstress Contractng among Rural Farmng Households n Malaw, Internatonal Journal of Development and Sustanablty, Vol. 3 No. 9, pp. 1862-1877. * Correspondng author. E-mal address: abdkhall@yahoo.com
Internatonal Journal of Development and Sustanablty Vol.3 No.9 (2014): 1862-1877 1. Introducton Land s one of the scarce natural resources that remans unequally dstrbuted n Malaw. Wth rapd populaton growth resultng n ncreased land fragmentaton, land holdng s becomng smaller that t s unable to sustan the households adequately. World Bank (2003) noted that majorty of the rural populaton produces 84% of total agrcultural output. Ths s from 1.8 to 2 mllon smallholder farmers who on average own only 1 hectare of land. The per capta land holdngs have declned from 1.53 hectares n 1968 to 0.8 hectares n 2000 (Government of Malaw, 2001). Estmates by the Malaw s Mnstry of Agrculture and Food Securty ndcate that 55% of households have an average land holdng sze of less than 1 hectare (Mnstry of Agrculture, 2003). Ths undermnes the desre to nvest on the land for mproved agrcultural ncome, and adopton of new farm technologes. The Mnstry advocates for a mnmum of at least 1.5 Ha of land to attan mnmum levels of sustenance throughout the year. One of the major constranng factors to ncrease agrcultural productvty and vablty s thus nsuffcent land requred for expanson. Despte the constrant, land plays a crtcal role n the lvelhoods of Malawans such that near landlessness has been lnked to poverty. Pressure on land s severe n densely populated areas of the southern regon of Malaw due to ncreased mgraton from the land scarce areas n Mulanje and Thyolo to less densely populated dstrcts of Mangoch and Mwanza. Land pressure s also becomng acute n the central regon where the captal cty s stuated (Greenwell, 2007).Wth growng populaton pressure on land, there s a need to develop an alternatve means to mprove land access n order to ncrease producton and productvty at household level. Land rental market would be one alternatve, whch s not well developed n Malaw. Land rental market s an mportant nsttuton n agrculture. Inter household resource exchange fostered n socal networks and assocaton membershp provdes an opportunty to combne labor durng farmland preparatons and harvest seasons among rural dwellers. Ths knd of resource sharng, to certan extent, reflects n plots of land n Malaw where the relatvely land rch households lend for free to the neghbor and/or relatves who need to produce food for home consumpton. In Malaw, land rentng and borrowng s practced but not legally recognzed (GoM, MGDS, 2005). Malaw s one of the countres where land market remans nactve. The lmted land market prevals tself n fxed rental form followed by borrowng contracts. Crop sharng, whch domnates the Eastern part of Afrca lke Ethopa, and sales market (common n Uganda) are scarcely observed or lackng n Malaw. Ths s manly due to hgher transacton costs, lmted household resources requred to contrbute n crop sharng contracts and/or legal prohbton n land sales as land use and access s closely montored by the vllage chefs n rural parts of the country (FGD report). The land rental market has got due attenton n agrculture based economes due to the ssues of equty, producton, and productvty assocated to t. Ths factor market s found to have promnent feature n transferrng land from less productve households constraned by soco-economc factors to more productve households as evdenced by Holden et. al., Dennger, and Ghebru (2007). Even wth major mperfectons n credts, labor, and nsurance markets, land rentng provdes asset of benefts to the poor people-by temporarly rentng out land they can get addtonal ncome wthout losng ther rghts to ther land. As ISDS www.sdsnet.com 1863
Internatonal Journal of Development and Sustanablty Vol.3 No.9 (2014): 1862-1877 stated by Ballestreso and Brescan (2008) that Transfer of land through rental provdes access to land to those wth hgh agrculture ablty but own lttle land or no land. Thus the land rental market allows for more effcent farm sze and provdes an opportunty for the landless to clmb up the agrcultural ladder. Moreover, as the off-farm economy develops, the rental market provdes mechansms for farm operatons wthout change n land ownershp. For the beneft of approprate polcy formulaton, knowledge n both dstrbuton of land, and determnants of land rental market would gude polcy makers n approxmatng mplcatons of skewed land holdngs on the economc performance across the agraran communty. It would also be the ratonale for redstrbuton or desgnng market orented land polcy regulatons whch facltate the opportunty to transfer land from land rch but less capable to the land poor and/or capable to cultvate households. Ths n turn has a welfare effect by reducng poverty at household level n partcular and the economy wde n general. Usng both Statstcal tools and Econometrc estmaton ths study amed at establshng mplcatons of the land market on households resource varaton. The study sought to address two key questons: Q1: Does land rentng mprove Food Securty of the land rental partcpants? It also hghlghts Q2: What are the determnants of land rentng n Malaw? Ths secton s followed by a bref background to the study whle secton three revews related lterature. The Methodology and Data requrements of the study are dscussed n secton four. Secton fve presents results and dscussons on the fndngs of the study and the last secton concludes the paper. 2. Background and revew Lke most sub-sahara Afrcan countres, Malaw s economy s hghly relant on agrculture whch accounts for about 90% of ts export earnngs and 45% of ts Gross Domestc Product. The sector employs about 85% of labor force (GoM, 2005). Ths explans the mportance of agrcultural land to the economy of ths country. However, t s constraned by lmted resources and rapdly expandng populaton. All Malawans rely upon access to land n one way or another. The country has a dversty of cultures, and major ethnc groups whch have ther own rules for the transfer of land to persons and generatons. For all groups, land s generally regarded as the man bass for socal securty (Gondwe and Moyo, 2008). In Malaw dfferent land allocaton systems have been developed to supply the populaton wth land. The basc objectve s to satsfy growng populaton wth enough space for food producton and shelter (Ercsson, 1999). The country s process of economc growth s one of contnuous structural transformaton channeled through varous lnkages between the ndvdual sectors of the domestc economy. Crucal for ths process and poverty allevaton are markets that operate effcently to accommodate decsons wthn and across households and sectors that lead to effcent use of nputs and outputs of agrculture. Accordng to the Government of Malaw, t allows all customary land to be regstered and protected by law aganst arbtrary converson to publc land. All customary land holders, defned to nclude entre 1864 ISDS www.sdsnet.com
Internatonal Journal of Development and Sustanablty Vol.3 No.9 (2014): 1862-1877 communtes, famles or ndvduals are encouraged to regster ther holdngs as prvate customary states wth land tenure rghts that preserve the advantages of customary ownershp and ensures securty of tenure. Accordng to Government of Malaw (2002) Prvate lease hold estates are created as subsdary nterests out of any prvate land, ncludng regstered customary estates wthout relnqushng the ownershp of the customary landholder. Ths provson allows tradtonal leaders, famly heads and ndvdual holders of regstered customary land to grant leases. That s, land under the customary tenure s consdered to belong to a vllage, and ndvduals n the communty who have the rght to cultvate the land and use t, and to dspose of t wth the lmts set up by the customary law of the trbe or clan. In ths case, therefore, the ndvdual does own the land wth lmted rght when comes t to exchange of land. The chefs, sub-chefs, and vllage head men protect the customary land aganst outsders (Nathole, 1985). Ths nsttutonal set up has lmted optmal land market that may shft land towards more productve households. 2.1. The Economcs of land rental markets The value of land s derved from resources exstng n/on t ncludng ts physcal structure, locaton, sze and qualty, and the summaton of ther ndvdual values calculated on monetary bass. Ths s often based on the future varatons of these components. In ths paper, the term "rent" should strctly mean the payment for the use of farm land. Davd Rcardo (1809) s credted wth the frst clear and comprehensve analyss of dfferental land rent and the assocated economc relatonshps. Rentng of land appears to be more wdespread and may be the most mportant form of land exchange. In Europe and Central Asa (ECA), land rentng has been expandng n terms of partcpaton and scale of partcpaton. For example, n Bulgara n 2003, only 3% of rural households had sold agrcultural land, whle 80% of them were rentng land n or out. Ad hoc evdence also ndcates that land rental plays an mportant role n the consoldaton of farms (Swnnen et al., 2006). Many lteratures confrm that rental markets can be an effectve nstrument to transfer land to the most effcent users and stmulate nvestment, f a number of condtons are fulflled, n partcular, suffcent tenure securty(skoufas, 1995; Holden et al., 2007; 2007; Tkabo et al., 2008; Ballesteros and Brescan, 2008; Holden et al., 2009). Among other factors, ths has to go wth the type of rental contracts used and the regulatons of the tenure system and lower transacton costs. In addton to effcency effects, rental markets may also have postve equty mpacts. The land rental market s also wdely practced n Phlppne agrculture. The study by Ballesteros et.al. (2008) descrbes rental actvtes n Phlppne agrculture whereby both sharecropped and fxed rent arrangements represent one fourth of the country s cultvated areas. The same study states that Share croppng s the preferred contract n all regons wth an average of 80% of total rented area under tenancy. It appears that corporate farms play a major role n the market. Where they domnate, rental market s actve and lease contacts are more common. Moreover, the study also dentfed that credt access and land endowment plays a role n determnng the probablty of land market partcpaton (bd, 2008). Revews of land market n Inda verfy that the market s affected by certan factors and found to be neffcent n adjustng the desred cultvable area by the household. Skoufas (1995) usng panel data from sx vllages n Inda establshed that hgher average female wage rate and large farm sze holdngs ISDS www.sdsnet.com 1865
Internatonal Journal of Development and Sustanablty Vol.3 No.9 (2014): 1862-1877 sgnfcantly reduce the probablty of leasng n land whle number of chldren n the household and hgher values of farmng mplements have negatve sgns on the supply sde of the market (bd). Studes have been conducted n dfferent countres of Afrca wth dfferent motves on the land rental markets, ncludng ts determnants, land reform and tenure securty and ther mplcaton for access to land, land nvestment, equty and effcency aspect of the actvty, mpact of certan nsttutonal changes on the market partcpaton, etc. For nstance, n the hghlands of Eretra, endowments of the non-land factors lke male labor force, oxen, and farm experence are found to be greatly affectng the probablty of partcpaton as well as the degree of partcpaton n the land rental market. Households poor n these factors tended to rent out land whle households rch n these non-land factors tended to rent n land (Tkabo and Holden, 2004). Shferaw et al., (2001) reports that land rental market n Ethopa ncreases effcences n creatng addtonal wealth f t contrbutes to use of more purchased nputs, mproved labor moblty/partcpaton n non-farm actvtes and hgh partcpaton n extenson package programs. Holden et al., (2010) reported that Effects on the allocatve effcency of the land rental market of the lowcost approach to land regstraton and certfcaton of restrcted property rghts mplemented n Tgray regon, Ethopa. The study found that low cost land certfcaton promoted partcpaton of female landlords to the rental market relatve to male counterparts. Ths was manly due to the fact that female headed households are constraned wth male labor force n order to tll the land and had been tenure nsecure because of ther lower barganng power before the offcal use rght certfcaton. The land market, n the form of sales and rental s also relevant n the Ugandan and Kenyan rural economes. In Uganda, Dnnger and Mpunga (2002), usng panel data (1999-2000) assessed the determnants of land rental partcpaton and systematc dfferences between sales and rental markets. The study dentfed, among others, that young and better educated households are more lkely to purchase land, and t was more dffcult for the landless, those wth few assets, and for households headed by wdows to acqure land through purchase markets n Uganda. Wth regards to the land rental, t was found that rental markets are more effectve than sales markets n transferrng land to larger households wth younger heads, thus provdng an opportunty to employ relatvely abundant famly labor (bd, 2002). Accordngly, t s much easer for landless households to gan access to land through rental than through sales markets. The report came up wth the fact that land rental market mproved productvty n a pro-poor way and helped to provde land access to those n need, especally landless households. On the other hand, based on cross-sectonal household data from two dfferent places of Kenya, Yamano et.al, (2005) establshed that those households wth greater number of women, educated, and asset wealth purchase land. In addton, number of oxen, and number of adult women are found to be postvely related to partcpaton n the market whereas the number of adult men s postvely related to the sze of rented n land. Ths study marked the mportance of land sales to those unable to nhert land though not used for mnor adjustment of landholdngs. As such, rental market s tended to be used by the partcpants for short-term adjustments n factor ratos. The paper ndcated the exstence of nverse farm sze- productvty relaton for whch t concludes as a base may be, potental landlords are reluctant to rent out land for fear of losng t. 1866 ISDS www.sdsnet.com
Internatonal Journal of Development and Sustanablty Vol.3 No.9 (2014): 1862-1877 Ths study focused on the mplcatons of the land market on household resource varaton n some dstrcts of Malaw. The study uses secondary household data collected durng 2008/2009. We verfy that ths study s dfferent n ts content and objectves from those that have been revewed. We are also not aware of any more studes so far n Malaw wth the same objectves, and therefore, clam that ths paper s our orgnal output by ts nature. 3. Methodology and data requrements 3.1. Theoretcal model Ths paper examned the land rental market partcpaton of households based on the theoretcal model developed by Blss and Stern (1982) and Skoufas (1995). These models have focused on the extent to whch land rented n or out by households adjusts to ensure that operatonal land holdng s adjusted towards the desred land area. Blss and Stern (1982) studed the land rental market n Planpur vllage n Inda usng a theoretcal model to assess whether the land adjustment n the rental market was complete as desred or less than desred due to transacton cost n the market. The model by Blss and Stern (1982) depcted that the ratonale of land rent can be explaned by mperfectons n the rural labor and credt markets. These mperfectons arse from ndvsblty of labor, lack of off-farm opportuntes and credt constrants n the agrcultural sector. Ths model found to be mportant as a theoretcal baselne n order to dentfy factors assocated to land markets operaton for adjustng the Desred Cultvated Area (DCA) n mperfect non land market stuatons among households. The functonal forms of the model are descrbed below. Expresson of the relatonshp s gven as follows; Assumng the absence of transacton cost n the land rental market, the household obtans the notonal demand for net land leased n, denoted by A *. However, n the presence of transacton cost the actual amount leased-n would be A. The functonal relatonshp s therefore specfed as n (). () A = h(a ) where: h s the adjustment functon whch s affected by the presence of transacton cost. Gven that A * s unobservable, t s assumed that households have a desred cultvated area (DCA) whch s assocated to agrcultural ablty,.e., famly labor (L), non land assets (A no ) and own land (A) sze. DCA s ncreasng n both L and A no (real value and tropcal lvestock unts). () A DCA Land endowment ISDS www.sdsnet.com 1867
Internatonal Journal of Development and Sustanablty Vol.3 No.9 (2014): 1862-1877 Af ( L, A no ) A Land leased n land leased out NLI DCA can depend on other varables asde from labor and other household assets. Lke off-farm employment, and household characterstcs (age, educaton, sex) and agro-ecologcal factors. Combnng () and () usng frst order Taylor seres expanson yelds the lnear equaton (): () no no A co h f1l h f2a h A co c1l c2a c3 * where, c o=constant term, h h / A, the slope of adjustment functon. And, f f and f f 1 no 2, whch mply margnal change n adjustment(dca) wth respect to A L labor and other household assets. If adjustment s done perfectly, the coeffcent of own land ( A ) s equal to one..e., f h ' 1 or c3 1, the actual cultvated area, A s equal to the DCA and the transactonal costs n the land rental market are nsgnfcant. On the other hand, f t s sgnfcantly dfferent from -1, t s a sgn of sgnfcant non-lnear transacton costs. Ths theoretcal postulate s based on the assumpton that both sdes of the market are opted to adjust operatonal holdng relatve to ther non land farm nputs. However, ths may not hold partcularly from the landlord sde n case when land rental s made n response to shocks n the absence of substtutes lke credt and nsurance markets. In such stuatons, household do rent out not for farm resource adjustment rather to overcome mmedate cash requrements under fxed contract arrangement wth more lkely lower barganng power to wards actual rental value. A 3.2. Hypotheses Based on the functonal forms presented above, two hypotheses were postulated: H 1: Better off households would choose to rent n land and may produce more output that would nfluence ther food reserve and agrcultural ncome, whle households poor n assets but own more land relatve to ther labor capacty would rent out ther excess land that they could not self-cultvate; Ths s based on revews n other countres that female headed households are constraned n famly labor, have less access to credts for mmedate cash needs due to poor socal relatons, and are vulnerable to dseases to self cultvate the land relatve to male headed households. H 2. There exst stronger reverse tenancy contract n rural households of Malaw. The hypothess s developed wth a presumpton that non land resource rch households are better off to employ/purchase farm nputs n order to cultvate extra land than ther counter parts who are captal constraned to do so. Hence, land rental drecton s from land rch but asset poor to wealther households. 1868 ISDS www.sdsnet.com
Internatonal Journal of Development and Sustanablty Vol.3 No.9 (2014): 1862-1877 Hypotheses H 1 and H 2 were tested usng bnary Probt (v) and censored Tobt (v) models respectvely. Bnary Probt (latent) model n the manner dscussed n (Green, 2003) s bult for land market partcpaton as: (v) A * o k 1 X where, o s constant, s are coeffcents, X s vector of explanatory varables, s the random dsturbance term. The probablty model for partcpaton s descrbed as; A = 1 f A * >0, Partcpatng as Tenant or Landlord 0 f A * <=0, No Partcpaton ether as tenant or landlord Models for both tenants and landlords are separately estmated aganst the non partcpants. The loglkelhood functon and ts dervaton can be obtaned and optmzaton can be done based on Green (Green, 2003). The explanatory varables used n each model are presented n appendx-1. The degree of partcpaton n terms of NLI would help to understand the extent of adjustment subsequent to partcpaton decson. Ths varable s consdered as postve for tenants but negatve for landlord households. Referrng to Wooldrdge (2009), we have derved Censored Tobt model from equaton (v) and t s set up as follows: (v) A max( 0, c L,A no o c L 1 c 2 A no c 4 R s c 5 S c6z h c A ) 3 Assumng L, R, A no, Z h 2, A ~ d (0, ) Unlke Lnear Probablty Model, Tobt model results n non-negatve predcted value for the dependent varable and have sensble partal effects for the range of explanatory varables (Madalla, 1983). That s why tobt (censored) model s preferred for ths estmaton Hypothess (H 3) s tested usng household assets (real asset values deflated usng 2006 prce, tropcal lvestock unts and famly labor) as proxes for wealth n the two models. 4. Results and dscussons Table 1 presents a comparatve assessment on dfferent assets among tenants, landlords, and nonpartcpants n the sx dstrcts. The assets consdered were own land sze (ownland), (NLI), tropcal lvestock unts (tlunts), household labor (hhlabour) total maze produce n 2009 (totmazpro09), and real value of assets (realvalue). Resource varatons among households across dstrcts were dentfed through statstcal tests of sgnfcance. ISDS www.sdsnet.com 1869
Internatonal Journal of Development and Sustanablty Vol.3 No.9 (2014): 1862-1877 Southern Table 1. Varatons n Resource Endowment among Households and Across Dstrcts Tenants Nonpartcpants Land lords Dstrct Varable Mean N se(mean) Mean N se(mean) mean N se(mean) Thyolo Own land sze(ha) 0.51 17 0.17 0.64 126 0.05 0.94 5 0.24 Net Land leased 0.16 17 0.037 0 126 0-0.15 5 0.03 Rent n land(1=yes0=no) 1 17 0 0 126 0 1 5 0 Tropcal lvestock unt 1.21 17 0.23 0.78 123 0.07 0.57 5 0.37 Household labor adult equva. 3.45 17 0.33 2.9 126 0.11 3.3 5 0.32 Total maze produce n 09 1135.3 17 207.1 978.5 126 111.5 496 5 127.1 Real valued Asset(1000Mkw) 11.5 17 8.45 5.02 126 1.98 1.21 5 0.74 Zomba Own land sze(ha) 0.79 25 0.14 0.89 204 0.037 0.83 10 0.12 Net Land leased(ha) 0.57 25 0.1 0 214 0-0.22 10 0.045 Rent n land(1=yes0=no) 1 26 0 0 214 0 1 10 0 Tropcal lvestock unt 1.95 26 0.37 1.28 210 0.1 0.67 10 0.14 Household labor adult equva. 3.01 26 0.2 2.98 210 0.09 2.49 10 0.43 Total maze produce n 09 1835.4 13 461.9 752.4 136 95.76 202 5 31.7 Real valued Asset(1000Mkw) 4.89 26 0.9 2.82 210 0.41 1.65 10 0.97 Chradzulu Own land sze(ha) 0.67 11 0.13 0.76 94 0.045 0.42 1. Net Land leased(ha) 0.18 11 0.03 0 98 0-0.32 1. Rent n land (1=yes0=no) 1 11 0 0 98 0 1 1. Tropcal lvestock unt 2.1 10 0.53 1.2 88 0.13 0.59 1. Household labor 3.91 11 0.42 3.23 94 0.14 2.8 1. Total maze produce n 09 960.5 11 180.9 575.9 94 43.17 400 1. Real valued Asset(1000Mkw) 0 11 0 0 94 0 0 1. Machnga Own land sze(ha) 0.69 13 0.21 1.25 132 0.05 1.32 13 0.15 Net Land leased(ha) 0.35 13 0.065 0 132 0-0.40 13 0.09 Rent n land(1=yes0=no) 1 13 0 0 132 0 1 13 0 Tropcal lvestock unt 2.16 13 0.99 1.87 132 0.24 1.51 13 0.76 Household labor 3.63 13 0.43 3.2 132 0.13 2.8 13 0.36 Total maze produce n 09 350 6 108.8 547.2 73 65.3 337.8 9 95 Real valued Asset(1000Mkw) 3.23 13 2.65 2.85 132 0.56 2.29 13 1.02 Central Kasungu Own land sze(ha) 1.13 27 0.10 2.08 294 0.12 1.78 20 0.24 Net Land leased(ha) 0.42 27 0.06 0 297 0-0.37 20 0.098 Rent n(1=yes0=no) 1 27 0 0 297 0 1 20 0 Tropcal lvestock unt 3.6 27 0.72 2.28 293 0.24 1.33 20 0.41 Household labor 3.53 27 0.3 3.58 297 0.09 3.63 20 0.34 totmazproduce09. 0.. 0.. 0. Real valued 7.54 27 1.74 5.27 297 0.64 4.97 20 1.46 1870 ISDS www.sdsnet.com
Internatonal Journal of Development and Sustanablty Vol.3 No.9 (2014): 1862-1877 Asset(1000Mkw) Llongwe Own land sze(ha) 0.86 24 0.13 1.043 161 0.05 0.92 11 0.2 Net Land leased(ha) 0.32 24 0.05 0 161 0-0.45 11 0.07 Rent n(1=yes0=no) 1 24 0 0 161 0 1 11 0 Tropcal lvestock unt 1.77 24 0.44 1.46 161 0.16 0.98 11 0.56 Household labor 3.27 24 0.2 3.05 161 0.08 2.6 11 0.29 totmazproduce09 1254.6 24 247.2 747.4 160 63.22 605 10 145.38 Real valued Asset(1000Mkw) 7.45 24 3.66 5.76 161 1.03 1.29 11 0.85 In Thyolo dstrct, t was noted that landlords are rcher n land whle average tropcal lvestock unts, real asset values, and household labor were sgnfcantly greater for tenants as compared to both landlords and non-partcpants. Non partcpants were poor n land and household labor relatve to both tenants and landlords. But they were found to be rch n lvestock and asset values relatve to landlords. In Zomba dstrct, tenants were wealther n labor, asset values, and tropcal lvestock unts but poorer n land than nonpartcpants. Tenants n Machnga dstrct owned sgnfcantly more land and non-land (lvestock, real assets, and household labor) resources. But stll some tenancy s takng place that could be for non-land resource adjustment. For Kasungu dstrct n the central regon, both landlords and non-partcpants have greater and sgnfcant land sze than tenants. However; tenants enjoy sgnfcantly hgher labor and lvestock unts relatve to landlords and non-partcpants. In Llongwe dstrct also of the central regon, non-partcpants have sgnfcantly hgher farm sze whle tenants enjoy more household labor. Consequently, t s perceved that non partcpants were expected to partcpate n ether sde of the market, especally n Zomba (poor n non-land), and Llongwe(less labor) dstrcts as landlords wth regards to ther land sze. Net land leased sze (NLI) vares across dstrcts. Ths mght be due to the varaton n the land endowments and non-land resources that lmt the leased land sze. Table 2 provdes an assessment of land market dstrbuton and concentraton across dstrcts. Ths helps to descrbe the cultural, geographcal and demographc pressure related to land rghts and on decsons to partcpate n land market. It was found that 22.3% are tenants rentng n whle 14.6% are landlords over all dstrcts. 10% of the sampled plots are ether rented n or out over the survey areas. Dstrct wse, We found more tenants n Kasungu(23%) followed by Zomba(22%). The dsparty may be due fear n some parts of Malaw where rentng out land s consdered as a sgnal that the owner need no more land and nablty to cultvate (socally ted to poverty). The study revealed that more landlords are located n Kasungu (33%) followed by Machnga (22%). Regon wse, the central regon dstrcts of Llongwe and Kasungu account for about 52% of the landlords whle the southern regon accounts for hgher percentage of the tenants (57%). It can therefore be deduced that rentng out s common n the central regon whle rentng n s common n the southern regon. Ths could b due to the fact that the central regon s characterzed by better nfrastructure and less populaton densty whereas the southern s populous and hence hgh pressure on farm land n whch case rentng out s not a feasble opton. Moreover, the land tenure securty ssue related to the nhertance and control over land system n the two regons contrbute to the extent of land market partcpaton. ISDS www.sdsnet.com 1871
Internatonal Journal of Development and Sustanablty Vol.3 No.9 (2014): 1862-1877 Table 2. Land market Partcpants across sx Dstrcts of Malaw from Plot Level Data Partcpants (%) Tyolo Zomba Chradzulu Machnga Kasungu Llongwe Total Tenants(%samples) 11.7 10.2 9.7 8.1 7.9 11.9 9.64 Tenants(of 118) 14.4 22 9.3 11 22.9 20.3 100 Landlord(%samples) 3.4 3.9 0.9 8.1 6 5.9 4.95 Landlords(out of 60) 8.3 16.7 1.7 21.7 33.3 18.3 100 The study also revealed that rental partcpaton vares across dstrcts n terms of the household resdental places wth respect to martal system,.e, whether patrlocal or matrlocal, Table 3. Table 3. Dstrbuton of Partcpants wth respect to Martal System and land ownershp. Dstrct Matrlocal Patrlocal Neolocal Total Tenants Landlord Tenant Landlord Tenant Landlord Tenant Landlord Tyolo 14 2 3 2 0 1 17 5 Zomba 16 6 9 3 1 0 26 9 Chradzulu 10 1 1 0 0 0 11 1 Machnga 4 10 3 3 6 0 13 13 Kasungu 3 4 23 16 0 0 26 20 Llongwe 11 3 13 9 0 0 24 11 Total 58 25 52 33 7 1 117 59 Rentng out s more common n patrlocal socetes, where around 56% of the landlords are located. Ths s related to the fact that land belongs to the husband sde and gves confdence on the securty of the plots n rentng out as compared to the matrlneal system of land possesson. Rentng n land s denser n Southern regon (57.3%) for couples resdng n the wfe s vllage (75.9%) as compared to the central regon whch accounts for 42.6% of the tenants and 69.2% from patrlocal socetes. Whereas the rentng out s hgher n the central regon (52.5%) for the couples lvng n the husband s vllage (75.8%). From the 1237 total plots surveyed, we found 9.5% of the plots were n fxed rental market, and 2.1% were borrowed. The knd of rental contract next to fxed rent and borrowng types was the land sales market (1.9%), Table 4. Ths verfes that the land sales market s the second best opton to fxed rental n exchange for cash. However, sales market data was dffcult to obtan as t s hghly assocated wth rsk of losng land due to the fact that sellng land s strctly forbdden. NLI (net land leased) s negatve for the landlords, zero for the non-partcpants and postve for the tenants, Fgure 1. Households are ordered by the sze of NLI, llustratng the relatve sze of the three categores of households relatve to the land sze rented n or out. The horzontal part of NLI represents 1872 ISDS www.sdsnet.com
Internatonal Journal of Development and Sustanablty Vol.3 No.9 (2014): 1862-1877 smallholders ratoned out due to hgher transacton costs or unwllng or/and already have enough land to partcpate n the market. Ths s an ndcatve to categorze households nto tenants, non-partcpants and landlords categores. The fracton of landlords s less than the tenants and far more less than nonpartcpants. Ths may demonstrate how nsecure t s to descrbe oneself as landlord than tenant. Table 4. Land Market Contracts n Malaw n 2008/2009 Season Contract Type Percent(%) of Total Plots Fxed Rent 9.50 Sales 1.86 Share Croppng 0.32 Borrowng 2.10 Total 13.78 6 4 2 0-2 -4 NLI (Ha) Fgure 1. Ranked households by Net Land Leased NLI for the tenants s postve and concentrated at a lower own land sze. Ths ponts out that the smaller own land sze drves to rent n land and hgher land ownershp forces landlords to rent out..e., the negatve NLI ncreases as the own land sze rses along the horzontal x-axs of Fgure 2 as can be observed from the downward slopng scatter plots. Ths may also ndcate that there are no economes of scale due to lumpness of some nputs. Ths s may be due to hoe-based cultvaton n Malaw (Lunduka. et al., 2006). One can observe that landless/land poor households could access land through rental market. The land rental market to some extent reallocates land from land-rch to land poor households leadng to more egaltaran land dstrbuton. Landless households are located vertcally perpendcular to the zero value. ISDS www.sdsnet.com 1873
Internatonal Journal of Development and Sustanablty Vol.3 No.9 (2014): 1862-1877 Fgure 2. The Net Land leased Vs. Own Land Sze The ratonale for land rental partcpaton by households n the country vares among the partcpants. It was dentfed (Table 5) that the majorty of the tenants (76%) have acknowledged that underszed own land relatve to other nputs compelled them to the market n order to ncrease ther operatonal holdngs. Majorty of the leased n plots were used for maze cultvaton, whle only 15% of tenants rented land to grow cash crops. Table 5. Reasons for land market partcpaton Reasons Tenants (%) Reasons Landlords (%) Increase land 76.1 Cash need 21.1 Grow cash crop 14.5 Assst other 38.6 Others 9.4 Has excess land 28.1 Others 12.3 One can examne that 60% of landlords seem to rent out land for mmedate cash needs that ether goes for assstance or home spendng. 28% of the landlords reported that they have excess land that they could rent out. Tenants had pad fxed rent of 30322MK per hectare on average for most plots they transacted, and average rental return was 7740MK per hectare for landlords. The per hectare value vares among dstrcts and even vllages due to, gven other factors, dfferences n land characterstcs, locaton and barganng power of the partners that may depend on the nstance of cash need. 1874 ISDS www.sdsnet.com
Internatonal Journal of Development and Sustanablty Vol.3 No.9 (2014): 1862-1877 5. Conclusons Land market n the form of fxed rental and sales contracts s emergng and growng n the country. Ths was however, found to be lmted n sharecroppng. Land and non-land asset varatons have derved rural households to enter nto land rental market. The momentarly transfer of land va the rental market s an mportant mechansm to allow the poor and land less to access agrcultural land, and obtan cash for landlords. The demand for land ncreases for households ownng more non-land assets. These factors have brought about strong reverse tenancy contract n the country. It was found that landless and land-poor Malawans relatve to ther famly labor rented n land. Fxed rental market s domnant and manly assocated to male labor force, household asset, tropcal lvestock unts, household health, and age varables. We found that rentng out of land may be a response to short-term stress when other alternatve sources of ncome have been depleted. Famly health shocks make land more mportant because the llness tends to narrow other possble lvelhood base and ultmately make land the only source to depend on. In general, ths paper suggests that land market partcpaton and degree of partcpaton n fxed rental form s found to be hghly assocated to land and non-land asset endowments, and household characterstcs. Ths market temporarly transfers land to more productve houses. Acknowledgments Ths artcle s based on the frst author s MSc thess from a jont program between Unversty of Malaw (Bunda College) and Norwegan Unversty of Lfe Scences. The frst author would lke to thank hs supervsors at both unverstes and s grateful to NORAD program for provdng scholarshp to pursue hs postgraduate studes. Comments receved from Rchard Kachule and Hastngs Moloko (both PhD canddates, Bunda, LUANAR) had greatly helped to re-organze and edt the orgnal artcle. References Ballesteros, M. and Brescan, F. (2008), Land Rental Market Actvty n Agraran Reform Areas: Evdence from the Phlppnes. Dscusson Paper Seres No.2008-26. Phlppne Insttute for Development Studes. Blss, C.J. and Stern, N.H. (1982), Planpur: The Economy of an Indan Vllage. Delh and Newyork: Oxford Unversty Press. Cross, S. (2002), Customary Land Tenure, Taxes and Servce Delvery n Rural Malaw: A Revew of Insttutonal Features of Rural Lvelhoods, LADDER Workng Paper No. 21. Dennger, K., Ayalew, A. Holden, S., and Zevenbergen, J. (2007), Rural land certfcaton n Ethopa: Process, Intal Impact, and Implcatons for other Afrcan Countres; World Bank, Washngton DC, USA, Norwegan Unversty of Lfe Scences, Aảs, Norway, Delft Unversty of Technology, Delft, Netherlands. ISDS www.sdsnet.com 1875
Internatonal Journal of Development and Sustanablty Vol.3 No.9 (2014): 1862-1877 Dennger, K. and Mpunga, P. (2002), Land Markets n Uganda: Incdence, Impact and Evoluton over tme, World Bank Dscusson Paper. Washngton DC: World Bank. Ercsson, A. (1999), Women s Access to Land. Proceedngs of UN-FIG Conference on Land. London, UK. Gondwe, I. and Moyo, G. (2008), Factors nfluencng sale and rentng out tradtonal land where Customary laws normally have regulated land transactons; unpublshed. Government of Malaw (2005), Malaw Growth and Development Strategy: From poverty to prosperty 2006-2011. Llongwe, Malalw. Green, W.H. (2003), Econometrc Analyss. New York: Macmllan. Holden, S.T., Dennger, K. and Ghebru, H. (2007), Impact of Land Certfcaton on Land Rental Market Partcpaton n Tgray Regon, Northern Ethopa; Norwegan Unversty of Lfe Scences, Norway. Holden, S.T., Otsuka, K. and Frank, M.P. (2009), The Emergence of Land Markets n Afrca. Impact on poverty, Equty and Effcency. Resources for the Future, Workng Paper. Washngton DC, USA. Lunduka R., Holden, T. and Oygard, R. (2006), Land rental market partcpaton and tenure securty n Malaw, Workng paper, Norwegan unversty of Lfe Scences, Aas, Norway. Mnstry of Agrculture, Irrgaton and Food Securty, Plannng Dvson (2003), Draft Concept Paper on the Establshment of the Food Securty Department, Llongwe, Malaw. Nathole, D.W. (1985). Land Tenure System and Agrcultural Producton n Malaw. Llongwe. Sadoulent, E. and de Janvry, A. (1995), Quanttatve Development Polcy Analyss. Baltmore: Johns Hopkns Unversty Press. Skoufas, E. (1995), Household Resources, Transacton Costs and Adjustment through Land Tenancy, Land Economcs, Vol. 71 No. 1, pp. 42-56. Smette, M. (2008), Rentng of Land-Study n two Vllages n Malaw, Workng paper. Norwegan Unversty of Lfe Scences, Norway. Swnnen, J., Vranken, L., and Stanley (2006), Emergng Challenges of Land Rental Markets n Europe and Central Asa, Workng paper. World Bank, Washngton, DC. Tkabo, M.O. and Holden, S.T. (2004), Land Tenancy and Effcency of Land use n the Hghlands of Ertrea. Norway: Norwegan Unversty of Lfe Scences, Norway. World Bank (2003), Afrca development ndcators: From the World Bank Afrca database. World Bank, Washngton DC. Yamano, T.K., Otsuka, F., Place, Y., Kjma and Nyoro, J. (2005), The 2004 REPEAT survey n Kenya (frst wave): Results. GRIPS Development data base No.1 Tokyo: Natonal Graduate Insttute for the Polcy studyng. 1876 ISDS www.sdsnet.com
Internatonal Journal of Development and Sustanablty Vol.3 No.9 (2014): 1862-1877 Appendx-1. Descrpton and Summary of Man Varables Used n the Analyss Descrpton of Varables (1a) Varable Name varable label Mean Dstrct 1=thyolo 2= Zomba 3= Chradzulu 4=Machnga 5=Kasungu 6=Llongwe Summary Statstcs Standard. Error Mn (1b) Max 3.74 1.6984 1 6 Regon 1=South 2=central 1.45 0.4976 1 2 Age Age Of The Household Member In Years 46.71 15.819 16 85 Schoolyears Number Of Years In School For The Household Member 5.31 3.970 0 24 Hghestclass Hghest Class Attaned By Household Member 4.67 3.556 0 15 Tmell Number Of Tmes Household Member Got Ill For more than two weeks n the last season 0.24 0.6975 0 4 Malehh Household Head Sex(1=Male, 0= Female) 0.78 0.4176 0 1 Consumer (Sum) Consumer Unts 3.99 1.56 0.8 9.2 Malelabour (Sum) Male Labour Force 1.76 1.040 0 5.5 Femalelabour (Sum) Female labour Force 1.45 0.756 0 4.2 Chldren (Sum) Chldren 2.7 1.61 0 7 Realvalue Deflated Real Values Of Assets Usng 2006 as base Year 4098 11868.14 0 144717 Tlunts Total Tropcal Lvestock Unts 1.6 2.64 0 17.2 totmazcon09 Total Maze Consumed In 2009 In Kg 803.6 783.76 12 8200 totmazpro09 Total Maze Produced In 2009 In Kg 797.8 990.67 20 7000 Market Dd You Rent In? 1=Yes 0=No 0.096 0.295 0 1 Market Dd You Rent Out Plot Last Year? 1=Yes, 0=No 0.049 0.2165 0 1 Plotdstance Plot Dstance From Home(M) 1167.7 2947.9 0 30000 Soltype General Sol Texture 1=Sandy 2=Loam 3=Clay 2.027 0.736 1 3 Slope Slope of The Plot 1=Flat 2=Slght 3=Clay 1.444 0.585 1 3 Plotfertlty Plot Fertlty 1=Very Fertle 2=Average 3=Not Fertle 2.052 0.626 1 3 NLI Net Land Leased(Ha) 169.165 1797.981-2 1.6 Ownland Owned Land Through Inhertance, Buyng &/Or Grant(Ha) 12117.79 12290.18 0 10.06 ISDS www.sdsnet.com 1877