Exploring long-term land improvements under land tenure insecurity

Transcription

1 Exploring long-erm land improvemens under land enure insecuriy Paper Presened a he New Zealand Agriculural and Resource Economics Sociey Conference Augus 2006 Sami Myyrä and Kyösi Pieola 2. MTT Agrifood Research Finland, Economic Research, currenly visiing a Waikao Managemen School, The Universiy of Waikao, Privae Bag 305, Hamilon. 2. MTT Agrifood Research Finland, Economic Research Summary This aricle explores long-erm land improvemen (lime applicaion) under land enure insecuriy on leased land. The dynamic opimisaion problem is solved by a sochasic dynamic programming rouine wih known parameers for one-period reurns and ransiion equaions. The model parameers represen Finnish soil qualiy and producion condiions. The farmer s decision rules are solved for alernaive likelihood scenarios over he coninuaion of he fixed erm lease conrac. The resuls sugges ha, as he probabiliy for non-renewal of he lease conrac increases, farmers quickly decrease invesmens in irreversible land improvemen and, hereafer, yields decline gradually. The esimaed decision rules are a par of larger se of farmer s decision rules o be aken care when land leasing and environmenal legislaion is renewed. Key Words Dynamic programming, land enure, land improvemens Inroducion The soils of Finland have been formed from acidic rock and ph values in agriculural soils in he counry are commonly below he recommended level. Therefore, liming is one of he basic amelioraive measures used o mainain good yields. There used o be a sligh bu seady gradual increase of soil ph from he 960's unil he 990's, bu paricularly during he las decade liming has been praciced far less han recommended. A he same ime srucural developmen has, in urn, gradually shifed culivaed land from land owners o enans. Over he years , he share of land culivaed under lease conracs by enans increased from 4.8% o 33.0%, i.e. by 28.2 percenage poins.

2 The sandard land lease conrac in Finland is a shor erm conrac wih fixed duraion and a fixed cash lease paymen per year. Abou 38% of all lease conracs have a duraion of five years. Wih only few excepions, he annual cash lease paymen is fixed per hecare of land when he conrac is signed. Conracs longer han 0 years are forbidden by law (Maanvuokralaki ). Tradiion in empirical sudies I is argued ha well-developed insiuional arrangemens and efficien asse markes should solve problems caused by land enure insecuriy (McConnell 983). Hence, economic lieraure analysing he implicaions of land enure insecuriy focuses for he mos par on developing counries where asse markes likely exhibi inefficiencies (Sjaasad and Bromley 997; Li e al. 998; Deininger 2003). In hese counries, land enure insecuriy has large implicaions, no only on land improvemens, bu also on household and sociey welfare. Even if he problems caused by land enure insecuriy are apparen, empirical sudies have no, however, been able o srongly idenify and characerise heir effecs. The daa have no been informaive enough, because in developing counries household access o farming inpus and markes can be severely resriced by insiuional, financial, and economic facors as well as by land enure insecuriy. Thus, i has been difficul o isolae he land insecuriy impac in saisical esing (Holden and Hailu 2002). Upcoming problem In richer counries, land enure insecuriy has no received much aenion in he economic lieraure. Even if land enure insecuriy is no ye a dominaing problem in hese counries, i may become such, paricularly in Less Favoured Areas (LFA), where producion coss are high and yields are low. The rend owards a more liberalised food marke and inernaionally harmonised agriculural policies decrease Marginal Value Producs (MVP) for agriculural inpus in LFA areas where farmer opions o adjus o hese rends are limied. Because he pay-back periods of irreversible land improvemens ge longer, we may expec ha in he LFA areas of norhern Europe land improvemens may decrease below he socially opimal levels if farmers are confroned wih significan land enure insecuriy. The decreased MVP canno alone explain he decrease in land improvemens, bu i is expeced o srenghen he implicaions relaed o land enure insecuriy, because enans, in paricular, may no longer have he incenives o make land improvemens in Finnish condiions as hey had in he pas. Tool for policymakers This aricle is o highligh he problem by solving and characerising he opimal decision rules o inves in long-erm land improvemens condiional on land enure insecuriy. Farmers opimal decision of liming is examined a he parcel level. The economic dynamic opimisaion model is numerically solved by dynamic programming wih known parameers for one-period reurns and ransiion equaions. The decision rules are solved condiional on alernaive scenarios on he likelihood ha he lease conrac expires. The

3 model parameers represen Finnish condiions and are based on field experimens. The resuls sugges ha opimal land improvemens decrease quickly as he uncerainy over conrac renewal increases and, hereafer, he yields decline gradually. Resuls confirm ha farmer s decision rules have o be aken care, when laws concerning environmenal resricions, land leasing and farmers access o he farmland are renewed. The model The Land Tenure Insecuriy Problem The farmer s problem is o opimise lime applicaions in a dynamic conex under cerain carry-over effecs. The opimisaion problem is normalised o one hecare of leased land and he crop used is spring barley. I is he mos common cereal grown in Finland. The model represens land ha has sufficien qualiy so ha he long-run equilibrium is o keep he soil ph saus a saisfacory levels from a sociey perspecive, i.e. wihou disorions caused by land enure insecuriy. In more marginal land areas, he land enure insecuriy problem is no he only insiuional facor ha is decreasing irreversible soil improvemens in a high cos counry wih a sparse rural populaion, such as occurs in Finland. The implicaion of land enure insecuriy was modelled hrough a sequence of fixed duraion (τ) land leasing conracs. We simulaed a five-year (τ=5) cash lease conrac, which is he sandard duraion for lease conracs in Finland. The average duraion of lease conracs in Finland is six years and 38% of all lease conracs have a duraion of five years (Myyrä 2004). Five-year conracs and commimens are ofen required in agri-environmenal programmes. Conracs longer han 0 years are forbidden by law (Maanvuokralaki ). The mos common lease paymen is a fixed cash paymen which is agreed when he conrac is signed. Under his conrac design, shor-erm land ren can be maximised simply by leasing he land for a farmer who pays he highes cash lease. The farmer has sure access o he land, unil he conrac s nex expiraion dae =nτ, for n=,2,3,,20. A each renewal dae =nτ, he renewal of he conrac is risky wih an exogenous probabiliy Prob nτ. Because coninuaion and expiry are muually exclusive, he probabiliy ha he lease conrac expires is (-Prob nτ ) and, once he conrac expires, he one-period reurns are assumed o say a zero forever. Thus, he expiry is assumed irreversible. The values of ph saus are divided ino seven classes as follows: poor, raher poor, fair, saisfacory, good, high and excessive (Viljavuuspalvelu Oy Soil Analysis Service Ld. 2000). The class limis are based on an exensive number of field experimens. Saisfacory is he arge class, considered sufficien for normal yields of field crops such as cereals and ley. Soil ph is deermined in waer suspension.

4 The Bellman equaion The opimisaion problem is formalised as a recursive finie horizon (T) dynamic programming problem ha is solved numerically by ieraing he Bellman s equaion (Bellman 957): { R (z,u ) + βprob [ V (z )] }, 0,,...,T V(z ) = maxu + + = () subjec o z + = g( z, u ) z given, T=00 where 0 Prob for all =nτ, and Prob = oherwise. where he opimal value funcion (V ) is he funcion of he curren sae vecor (z ); u is he conrol (also called he decision rule or policy funcion); R (.) is he one-period ne reurn funcion; β is he discoun facor. The opimal value funcion is consrained by ransiion equaions, in which g(.) is a funcion. I deermines he connecion beween he curren sae and conrol and he nex period sae. The opimal soluion is a funcion of he iniial sae z. This specificaion generalises he models presened in Kennedy (986). The one-period ne reurn (R ) is he difference beween he one-period revenue from selling he yield minus he expendiure of purchasing he conrol. Because oher facors are held as fixed in he analysis, hey can be suppressed and he one-period reurn funcion is: R ( z, u ) = p y( x, u ) w u (2) where y(x,u ) is a deerminisic yield response funcion, specified separaely for phosphorus and lime (see below). The las erm w u is he expendiure on inpus, i.e. lime and phosphorus feriliser. The ime horizon (T) was se a 00 years which guaraneed ha no only he decision rules bu also he sock variables and he value funcions converged. The sae vecor (z ) consiss of he ph (x ) and he oupu prices (p ) as well as he price of he conrol variable (w ). Prices are assumed deerminisic such ha he curren prices prevail in he fuure. Thus, he ransiion equaions for prices are simply p = + = p+ = p and w = w w (3) Funcional forms As far as soil ph is concerned, here are pleny of repors abou he negaive impacs of excessive ph values on crop growh, mosly caused by deficiency of meallic micronuriens (Cu, Zn, Mn, Fe). These adverse effecs occur in neural and basic soils wih ph values above 7, which level is irrelevan in our sudy. Research resuls

5 unanimously sugges ha liming acidic soils increases gives a posiive yield response and gradually, when approaching ph 7, a plaeau is reached (see Lahwell and Reid, 984). We operae exclusively in acidic soils, which have lower han opimal ph for he crops (see Havlin e al., 2005). Therefore, he Mischerlich equaion, no accouning for he decrease of he yield upon excessive inpus, is well reasoned in our sudy. Lime (L): yield response and carry-over Liming (u l ) is measured in onnes per hecare (/ha) and he sock of lime (x l ) is measured in soil ph. Liming has only an indirec effec on yield via soil ph. The yield response o he soil ph is described by a Mischerlich funcion (Kempainen e al. 993; Myyrä e al. 2003) y( x, u ) = e l l 3.85 ph (4) The ransiion equaion describing he carry-over effec and he effecs of liming o he soil ph is (Kemppainen 993) x l = ph + = 0.049u + ph 0.05 l + (5) Wih no liming, he average annual decay rae in he soil ph is 0.05 ph-poins, which implies ha annual amoun of liming required for mainaining he exising ph level is on average, 0.3 onnes per hecare. The iniial sae was se a x l 0 = ph 0 = 5.8, which was esimaed o be an average on leased plos in cenral and norhern pars of Finland (Myyrä e al. 2003). The prices and he scope of he model are summarised in Table. Table : Prices and he scope of he opimisaion problem. Price of barley a Price of lime a if applied -3 onnes/ha if applied 4-5 onnes/ha Discoun facor (β) Time horizon (T) Duraion of single conrac (τ) 0 /onne /onne 33.6 /onne / years 5 years a The prices are a he farm gae. The price of liming includes field spreading. The spreading incurs an exra cos a low applicaion raes. Resuls Liming is o some exen lumpy as i is expensive o disribue small amouns of lime (see he price hresholds in Table ). Therefore, if he iniial soil ph (5.8) is wihin he

6 biological arge range, which is he case in he simulaions, land enure insecuriy does no make a difference in he opimal liming rules (Figure ). I does no pay o apply lime on land wih a ph level exceeding 5.4, excep when farmer access o land cerainly coninues, eiher hrough repeaed conrac renewals or land ownership. In his case, he decision rule converges wih he opimal behaviour around he long-run equilibrium wihou land enure insecuriy, which is also opimal for sociey. When he iniial ph level decreases 2 and soil acidiy increases, decisions as o lime vary according o he uncerainy over he lease renewal (Figure ). If he conrac is going o expire wih cerainy, i pays o lime a he beginning of a five-year conrac if, and only if, he soil ph is below an exremely low value of 5. If he odds are slighly in favour of conrac renewal (Prob=0.6), i pays o lime a he beginning of he five-year conrac if he soil ph is below 5.2. Figure : Developmen of soil ph in a sequence of 5-year lease conracs, condiional on alernaive conrac renewal probabiliies (Prob). The upward sloping jumps in soil ph indicae poins where lime is applied. The downward sloping line races depleion of he soil ph when lime is no applied 6 5,75 ph 5,5 5,25 5 4, ime Prob= Prob=0.6 Prob=0 The risk free long-run equilibrium is o mainain soil ph above 5.4, represening a fair soil ph saus in he soil ype sudied. When conrac renewal is likely, e.g. Prob>0.5, i 2 The soils of Finland have been formed from acidic rock and ph values in agriculural soils are decreasing because he use of nirogen ferilisers and acidic rain.

7 is sill advanageous o mainain he soil ph above 5.2. Opimal iming is o apply liming immediaely afer he new conrac is signed. Neverheless, when he likelihood for conrac renewal decreases and he odds are in favour of conrac erminaion (Prob<0.5), he soil ph is allowed o decrease below 5.2, which mos commonly represens he feriliy class raher poor or poor 3. If liming is negleced, i resuls in a gradually decreasing soil ph which will evenually decrease yields (ceeris paribus). Figure 2 races ou paerns of consequen five-year conracs summing o a 00-year period. When he land has been culivaed under several subsequen five-year lease conracs and he odds of coninuiy of he conrac has been expeced o be 0, an lease holders opimal policy leads o he yield of around 3,300 kg/ha. The yield decrease from he 3,670 kg/ha seady sae equilibrium wihou land enure insecuriy is 230 kg/ha (in relaive erms, 6 per cen). If he odds are slighly in favour of conrac renewal (Prob=0.6) he seady sae yield will be decreased only 20 kg/ha (3%) o 3,550 kg/ha. Figure 2 : Prediced yield response o opimal liming under alernaive probabiliies (Prob) for a renewal of each five-year lease conrac (ceeris paribus). yield esimae kg/ha ime Prob= Prob=0.6 Prob=0 3 The class limis depend on he paricle size disribuion and he organic maer conen of he soil.

8 Discussion The resuls sugges ha he opimal decision rules on irreversible land improvemens wih long pay-back periods vary subsanially according o he exen of land enure insecuriy. Land improvemens decrease below he social opimum when he likelihood for conrac renewal decreases. Therefore, he curren endency of gradually increasing share of land culivaed under simple fixed duraion cash lease conracs poses a problem in mainaining land improvemens and soil feriliy ha are sufficien for maximising social welfare. This will finally urn ino decreased yields and weakened food supply. Furher, he incenive problems caused by land enure insecuriy will hamper he efficiency of environmenal programmes o decrease nurien run-offs since he sandard is ha also hese programmes require irreversible land invesmens wih long pay-back periods. Resuls also claims ha required five-year conracs and commimens in agrienvironmenal programmes are oo shor o give real incenives o enan for yield improving invesmens in leased plos. The resuls provide srong signals, and indeed jusified concerns, o expec ha land enure insecuriy will cause in he fuure more severe problems o land improvemens han are currenly being revealed by he saisics on soil ph (Myyrä e al. 2005). The subsanial implicaions, as suggesed by our opimisaion model, are suppored by aggregae marke behaviour even if his behaviour canno, due o daa limiaions, be direcly linked o land enancy and uncerainy over conrac renewals. The demand for lime has been decreasing rapidly as he share of land culivaed under lease conracs has been increasing. I has o be noed, however, ha marke rends have been affeced also by oher insiuional and economic facors, such as he decreasing marginal value produc for lime. Our resuls hold only o simple cash lease conracs where he likelihood of conrac renewal is exogenous. The resuls do no generalise repeaed dynamic games in which he farmer s repuaion may have significan implicaions on opimal decision rules. Thus, our analysis does no accoun for he possibiliy ha land improvemens may be used o increase he likelihood of conrac renewal (Sjaasad and Bromley 997). One of he main goals of he Common Agriculural Policy (CAP) in EU is o increase produciviy in agriculure. The resuls signal ha reaching his goal may be seriously hampered by gradually increasing land leasing unless he land enure insecuriy problem can be solved by beer conrac design. These conrac designs mus include he feedback from land feriliy o land value, however one can claim ha land prices are no longer driven by agriculural facors. One opion o lower he uncerainy caused by fixed duraion cash lease conracs would be o relax he legislaive regulaions on he maximum duraion of land lease conracs. The conclusion of his paper is ha farmers (owners and enans) may follow he socially opimal pah in liming, if awareness of changes in land feriliy is absolue, land produciviy is valuaed correcly and land lease markes are perfec. However, hese condiions are no reached in simply cash lease conrac markes.

9 References Bellman, R., 957. Dynamic Programming. New Jersey. Deininger, K., Land Policies for Growh and Povery Reducion. A World Bank Policy Research Repor. Havlin, J.L., Beaon, J.D., Tisdale, S.L., Nelson, W.L., Soil Feriliy and ferilizers. 7 h Ed. Pearson Prenice Hall, New Jersey, USA. Holden, S., Hailu, Y., Land redisribuion, enure insecuriy and inensiy of producion: A Sudy of farm household in souhern Ehiopia. Land Economics, 78: Kemppainen, E., Jaakkola, A., Elonen, P., 993. Pelomaiden kalkiusarve ja kalkiuksen vaikuus viljan ja nurmen saoon. In Finnish, wih an English absrac: Effec of liming on yield of cereals and grass. Maaalouden ukimuskeskuksen iedoe 5/93. Kennedy, J., 986. Dynamic programming. Applicaions o agriculure and naural resources. Ausralia. Lahwell, D.J., Reid, W.S., 984. Crop response o lime in he norheasern Unied Saes, pp In: Adams, F. (Ed.). Soil Acidiy and Liming. Agronomy Monographs 2, 2 nd Ed. American Sociey of Agronomy, Crop Science Sociey of America and Soil Science Sociey of America, Madison, Wisconsin, USA. Li, G., Rozell, S., Brand, L., 998. Tenure, land righs, and farmer invesmen incenives in China. Agriculural Economics, 9: Maanvuokralaki. (Law of land leasing) Referred hp://finlex.edia.fi. McConnell, K., 983. An Economic Model of Soil Conservaion. American Journal of Agriculural Economics, 65: Myyrä, S., Keoja, E., Yli-Halla, M., Pellon hallinaoikeuden yheys maanparannuksiin - esimerkkinä kalkius ja fosforilannoius. In Finnish, wih an English absrac: Land enure and land improvemens in Finland - he case of liming and phosphae ferilisaion. MTT:n selviyksiä 37. Myyrä, S., Keoja, E., Yli-Halla, M., Pieola, K., Land Improvemens under Land Tenure Insecuriy - he Case of ph and Phosphae in Finland. Land Economics, 8: Sjaasad, E., Bromley, D., 997. Indigenous Land Righs in Sub-Saharan Africa: Appropriaion, Securiy and Invesmen Demand. World Developmen, 25: Viljavuuspalvelu Viljavuusukimuksen ulkina peloviljelyssä. In Finnish. Tile in English: Inerpreaion of soil es resuls in field crop producion. Yearbook of farm saisics Maa- ja mesäalousminiseriön ieopalvelukeskus, Helsinki.