Prce Responsveness of Salmon Supply n the Short and Long Run Krstn Roll, Trude Thomassen and Sgbjørn Tveterås Farmng the Sea, Refsnes 2007
Introducton Productvty growth and compettveness among salmon producers ndcates a responsve salmon supply n the long run Several studes have documented salmon productvty growth and other aspects of supply Salvanes (1989), sche & Tveterås (1999), Tveterås (1999), Guttormsen (2002 Tveterås and Heshmat (2002), Kumbhakar (2002), Tveterås and Battese (2006) However, no study have nvestgated the supply responsveness to prce changes of salmon producers, whch s surprsng gven the cyclcal behavour of salmon supply Steen, sche and Salvanes (1997)
Introducton The ndustry has experenced prolonged perods of low and hgh profts, whch ndcate that salmon supply s not easly adjusted n the short run In ths paper we try to shed some lght on the cyclcal behavour of the ndustry by usng a translog proft functon to estmate short- and long-run supply elastctes for salmon
Norwegan export prce, unt cost and unt margn 1985-2006 (2006=1) 100 80 60 Unt margn Prce Cost 25 20 15 NOK/kg 85 87 89 91 93 95 97 99 40 20 0-20 -40 2001 3 5 10 5 0-5 -10 *Estmate
Background lthough the productvty growth across the ndustry mples a strong supply response to ncreased competton, t s unclear how responsve supply s to prces n the short and n the long run 1. Because of the bologcal producton cycle there s a substantal lag before producers are able to respond 2. Regulatons on producton volumes, nputs, locaton etc. nfluence the salmon supplers ablty to respond 3. It takes tme to adjust producton capacty because of quas-fxed captal 4. Producton rsk such as dsease outbreaks can msalgn demand and supply
1. The bologcal producton process Cyclcal proftablty of the salmon ndustry has been assocated wth characterstcs of the bologcal producton process Salmon producton s carred out n two phases; The frst s a fresh water phase where the ova are hatched and the smolts reared n tanks or n cages n fresh water lakes. Ths takes around 11-12 months. The second phase nvolves growng the smolts n cages n sea water untl they reach market se, whch may take around 10-18 months.
2. Regulatons affectng salmon supply Regulatons n Norway Farmng concesson (lcense) 12 000 m 3 maxmum bomass of 780 (900) tonnes Locaton of farm ste Cage volume Fsh densty (from 1991) maxmum densty of fsh s 65kg/m3 Feed quota (1995) Types of regulatons vares, but salmon producers n all countres faces some degree of regulaton that affects producton possbltes
3. Captal Investments Producton and marketng of farmed salmon are assocated wth sunk cost n the form of nvestment n educaton and tranng of personnel, captal equpment, market research and advertsement. Snce level of nvestment s chosen condtonal on nformaton before producton begn, t may later turn out to be suboptmal compared to the realsed output level and market prces.
Model We estmate a restrcted proft functon where some quas-fxed nput factors are fxed n the short run McFadden (1978), Lau (1976), Brown and Chrstensen (1981), Kulatlaka (1985) Squres (1987) The proft functon s restrcted (or partal) because frms are assumed to be n a statc equlbrum wth respect to a subset consstng of the varable nputs and output statc because future levels of quas-fxed factors are not affected by past departure from long-run equlbrum
Model The short-run (restrcted) and long-run proft functons descrbe the technology and costs n the short- and the long-run equlbrums. The restrcted proft functon: HR(P; Z) P : prces (varable nputs and outputs) Z : levels of quas-fxed factors The long-run total proft functon: HT(P, P Z ) = HR(P, Z*(P, P Z )) P Z Z*(P, P Z ) P* Z : shadow prces (equal to the market rental prces P Z of the quas-fxed by the envelope condton)
Short- and long run average cost
Translog restrcted proft functon ln HR = r D r r ln P j j ln P ln P j T T ln Z TT T 2 ln Z 2 T ln P ln Z ln PT ln T ZT The condtonal revenue cost shares obtaned by Hotellng s lemma are ln HR ln P = PY HR = S = j j ln P j ln Z T T
Long-run elastctes. 2 2 ( S S ) ( SS ) E = 2 S S ( S S ) ( j SS j ) ( SS )( j S js ) Ej = 2 S S ( S S ) E = 2 S ( S S ) E E S ( S S ) = 2 S ( S S ) ( S S ) = 2 ( S S )
Data nnual data from 1985-2004 Source: Norwegan Drectorate of Fsheres Frm level data on Revenue Costs Output Input Regon
Emprcal results Short-run elastctes Output Feed Labour Output 0.0484-0.048-0.001 Feed 0.516-0.487-0.029 Labour 1.618-0.672-0.946 Long-run elastctes Output Feed Labour Captal Output 1.415-0.767-0.080-0.568 Feed 1.827-1.191-0.098-0.538 Labour 2.943-1.378-1.024-0.541 Captal 5.810-1.575-1.201 3.034
Concludng remarks Salmon producers have lmted possbltes to respond to prce changes n the short run Changng feed volume or formulatons may gve some room for adjustments n producton In the long run the ndustry s prce responsve The combnaton of hgh responsveness n the long run and lmted responsveness n the short run s most lkely the chef reason for the pronounced cycles n the ndustry s long as the ndustry reman compettve wth many producers the cycles wll most lkely reman