Research on Growth Drag of Water Resource on Agricultural Development in China

Transcription

1 Research on Growth Drag of Water Resource on Agrcultural Development n Chna Xueyuan Wang College of Economy, Zhejang Gongshang Unversty, Hangzhou, Chna Emal: wxyrocky@163.com Contrbuted Paper prepared for presentaton at the Internatonal Assocaton of Agrcultural Economsts Conference, Bejng, Chna, August 16-22, 2009 Copyrght 2009 by [Xueyuan Wang]. All rghts reserved. Readers may make verbatm copes of ths document for non-commercal purposes by any means, provded that ths copyrght notce appears on all such copes.

2 Research on Growth Drag of Water Resource on Agrcultural Development n Chna Abstract Based on the growth drag conceptual model of Romer, ths paper computes the growth drag of water resource on agrcultural development through usng panel data(1997~2006) of Chna s all provnces. The result shows the growth rate of agrcultural producton value fell % per year on average because of the shortage of water resource. The number s seemngly small, but accordng to calculaton, under the drag of water resource, the rato wll be reduced by 2.66% n Furthermore, t wll be 3.74% n 2040 and 4.82% n Keywords- water resource; agrcultural producton; growth drag INTRODUCTION As s known to all, water s necessary basc element for crop. Wthout water, there s no agrculture. Snce the total quantty of water supply s nsuffcent, together wth ndustral and urban ncreasng demand, the current stuaton of agrcultural water s qute serous n Chna. Durng the perod of the Tenth Fve-Year Plan, rrgaton water s shortage of 150 bllon m 3, and about 0.26 mllon hm 2 farmlands suffer from drought, whch lead to decrease by 35 bllon kg grans per year [1]. Havng wrtten an artcle enttled "Who wll feed Chna", Brown Lecestershre, the drector of U.S. World Observaton Insttute, shocks the world. But, n 1998, he publshed another artcle "Chna's water shortage wll shake the world food securty" whch proved Chna's most serous resource crss was not arable land, but water. Based on the future analyss and forecast of Chna's water demand, Brown ponted out Chna's agrculture water was facng wth ferce competton wth urban and ndustral water demand. Wth the depleton of ground water and worsenng water polluton, water shortage wll greatly threat Chna's agrcultural development, even the food securty of the world [2]. Is the stuaton really serous as the forecast? What s the truth? Scentfc answer needs the support of strong evdences, partcularly quanttatve analyss n lne wth actual stuaton. The author, wth the conceptual model of Davd Romer s "growth drag ", constructs smlar model to measure the extent of constrants of Chna's water resource to agrcultural development, so as to provde references for polcy-makers and related researchers. THE DEFINITION OF GROWTH DRAG New growth theory tells us economc development s sgnfcantly related to resource and land.

3 Compared the case of constrants wth non-constrants, the growth rate of economy s greatly dfferent, whch stems from the "growth drag" of lmted resources [3]. Xue Junbo [4], Xe Shulng [5] and Yangyang [6], usng specal model, respectvely assessed the "growth drag" of Chna s sol and water resources. Though they all translated "growth drag" as "growth tal effect," the meanngs are markedly dfferent. Accordng to Xue, "growth tal effect" should be defned as the extent of reduced economc growth rate under resource constrants. Whle Xe frmly nssts on, for any country and regon, the process of economc development wll nevtably lead to the consumpton of resources. Furthermore, ntal nput wll affect the latter growth because of lmted resources, whch s called "growth tal effect". Based on the above mentoned analyss, Yang Yang beleves that "tal effect" generally refers to the lag effect of whch wll contnue to play an mportant role n the next stage. In fact, the concept should be translated as "dampng growth". In terms of land resource, ts value equals to the dfference of economc growth rate between exsted and no-exsted constrants. Clearly, for the connotaton of "growth tal effect", Xue and Yang s smlar to Roemer s "growth drag", but vastly dfferent from the defnton of Xe. The author also agrees wth Yang-Yang's vew of whch "tal effect" s not sutable to descrbe the "growth drag" of economc development restrcted by resource and land. Accordng to Roemer s senor macroeconomcs (second edton), t seems more reasonable to translate "growth drag" as "growth resstance". The purpose of ths study s to explore the mpact of water resource constrants on agrcultural producton. Referred to Roemer s defnton, the "growth resstance" of water resource should mean the growth dfference between "there are no water restrctons" and " water restrctons exst" n the process of agrcultural development. MODEL AND METHOD Assumpton Revson In order to assess the extent of resource constrants on economc growth, Roemer constructed a C-D producton functon model ncludng natural resource and land, the form s as followed: Yt ( )= Kt ( ) Rt ( ) Tt ( )[ AtLt ( ) ( )] α β γ (1 α β γ) α> 0, β> 0, γ> 0, α+β+γ< 1 (1) Among them, Y s output, K s captal, R s natural resource, T s land, L s labor, A s knowledge or effectve labor. When exstng restrctons, Tt () = 0and R() t = -br() t (b>0). When not exstng restrctons, Tt () = ntt () and R() t = nr() t These mean land and resource both grow together wth

4 populaton. The growth drag s: Drag = βb + ( β + γ) n 1 α (2) That s to say, the drag s constantly ncremental along wth resource share β, land share γ, the utlzaton of declnng resource b, the growth rate of populaton n and ncreasng captal share α. In order to see more clearly the nfluence of water resource to the "growth drag" of Chna's agrculture, the author wll adjust and amend the model n accordance wth the facts. Compared wth 1997, total water resource decreases by bllon m 3 n 2006; agrcultural water consumpton decreases by bllon m 3 ; water share also drops by 7.2 percentage ponts and rrgaton water per acre s approxmately reduced 50 m 3. The avalable water resource of Chna's agrculture s reducng gradually because of lmted water supply as well as ncreasng water demand from ndustry and cty, whch leads to the declne of rrgaton water per acre and hnders agrcultural producton. In order to more accurately reflect the mpact, the author wll quanttatvely calculate the degree of constrant of water resources to the growth of agrcultural output Crop areas Agrcultural water Year Crop areas Agrcultural water Fgure 1 The change trend of Chna's agrcultural water and crop areas (1997 ~ 2006) Data sources: Statstcal Yearbook of Chna's rural areas ( 1998 ~ 2007 ) and Chna's Water Resource Communque(1998~2007), calculated and arranged by the author. As s shown by Fgure 1, from 1997 to 2006, agrcultural water doesn t change at the same proporton as crop areas. Therefore, when analyzng, resource should not be supposed to a fxed constant or the symbol of growth rate s dentfed as negatve or postve beforehand. Based on prevous studes, the author makes followng revsons to the model. When exstng restrctons of water,

5 W () t = g W () t. Otherwse, let W () t = g W () t.ths means the change velocty of agrcultural W T water and crop areas s dfferent. Here, s the growth rate of crop areas n regon. g W s the growth rate of agrcultural water n regon, g T Model Constructon follows: Now, we stll use the form of C-D producton functon. But the model constructon s changed as Y()= t K () t W () t T()[ t A() t L()] t Y ( t ) : Agrculture outputs of regon at the tme t ; α β γ (1 α β γ) K ( t ) : Captal nvestment of regon at the tme t ; W () t : Agrculture water nputs of regon at the tmet ; T ( t ) : Land nputs of regon at the tmet ; L () t : Labor nputs of regon at the tmet ; A () t : The effectveness of knowledge or work of regon at the tme t ; α β γ: Parameter, and α>0, β>0, γ>0, α+β+γ<1 (3) Takng the logarthm on both sdes of the formula (3), we can obtan the followng equalty: ln Y( t) = α ln K( t) + β ln W( t) + γln T( t) + (1 α β γ)[ln A( t) + ln L( t)] (4) Further deducng, we can get: g = αg + βg + γg + (1 α β γ)( g + g ) (5) Y K W T L Among them, g, g, g, g, Y K W T g L and g respectvely represent the growth rate of Yt () K () t Wt () Tt () L () t A () t. Accordng to Roemer, f locatng n a balanced growth path, g Y s certanly equal to g K. Put t nto the type (5) and solve the equaton. g (1 α β γ)( g + g ) + γg + βg = 1 α bgp L T W Y (6) bgp g Y : growth rate of agrcultural output n a balanced growth path.

6 It also means the growth rate of agrcultural output per acre s: bgp bgp bgp (1 αβγg )( + gl) + βg W (1 γg ) g(/) YT= gy gt = 1 α T (7) After deducng the growth rate bgp g( Y / T ) constrants and get the followng output growth rate per acre., wth the same method, we study the stuaton of no-exstng bgp (1 α β γ)( g + gl) + ( γ+ β 1) gt g ( Y / T) = (8) 1 α Therefore, durng the process of agrculture development, the "growth drag" of water resource s: bgp bgp β( gt gw ) Drag = g ( Y / T) g( Y / T) = (9) 1 α From the equalty (9), we can see the "growth drag" wll ascend along wth ncreasng captal output elastcty(α), output elastcty of water resource (β) and crop areas. Ths ndcates: although captal, land and water resource are man nput elements of agrcultural producton, agrcultural sustanable development can not excessvely depend on them. Thus, the only way to reduce "growth drag" s to promote technologcal progress whch rases tself status n agrcultural producton. In addton, under CD producton functon, the substtuton elastcty of producton elements s a fxed constant 1. In fact, the elastcty may not be equal to 1. When t s less than 1, as growng scarcty elements, the contrbuton share of nput to output wll ncrease and become more mportant. On the contrary, the share wll drop [7]. MEASURES OF GROWTH DRAG Data Descrpton The research's goal s to measure the constrant extent of water resource shortage on agrcultural producton. Consderng avalable data, we adopt panel data set of Chna's all provnces and autonomous regons from 1997 to All data are derved from "Statstcal Yearbook of Chna's Rural Areas" and "Communque of Chna's Water Resources". The output varable, Y () t, s reflected wth agrcultural output. As numercal value s from current year's prce, we wll take 1996 as the base year and adjust Y () t accordng to comparable prce. Referrng to past research lteratures [8], the artcle uses employment fgure of agrculture, forestry, anmal husbandry and fshery as. Lt () K() t

7 s measured by the total power data of agrcultural machnery. Wth regard to, because output data manly am at narrow plantng agrculture, here, we adopt rrgaton water data. On average, t accounts for 90% [9] of total agrcultural water. Therefore, for the data of publshed Yearbook, the author converts them through multplyng by 90%. In addton, the mssng data are polshed by dfferental technque. Quanttatve Analyss Accordng to the formula (9), αand β, parameters of C-D producton functon model, are bass of quanttatve estmatng the "growth drag". The paragraph frst carres on the parameter estmaton to multple regresson model of formula (4) through usng the software STATA10.0. Table 1 shows the W () t results of parameter estmaton. As can be seen, the value of 2 χ s of 4.79 on Hausman test, whch ndcates the estmaton results of fxed effects (FE) and random effects (RE) model dffer just slghtly accordng to the panel data set. Moreover, the results of F-test and Breusch-Pagan LM random effects test also show the above two methods are superor to ordnary least squares (OLS), more than 96% n goodness of ft. However, further tests dsplay there exst frst order autocorrelaton, relevant secton and heteroscedastcty n both fxed effect model and random effects model. Table 1 Estmaton results of fxed effects and random effects model Independent varable Coeffcent fxed effects model random effects model lnt ***(0.0346) ***(0.0327) lnl ( ) (0.0437) lnk ***(0.0225) *** (0.0199) lnw ***(0.0454) *** ( ) constant ***(0.5124) *** (0.2226) Goodness-of-ft: Wthn Between Overall Sample number F test:u =0 F(30, 275)=71.46 Breusch-Pagan LM: Random effects test χ 2 = Wooldrdge Autocorrelaton test F= LM= Breusch-Pagan LM: Independent testng χ 2 = Wald Heteroscedastcty test χ 2 = Hausman test χ 2 = 4.79, Prob>χ 2 = Note: 1 Agrcultural output value s seen as dependent varable 2 *ndcates a 10% sgnfcant level, ** ndcates a 5% sgnfcant level and *** ndcates a 1% sgnfcant level. Standard errors n parentheses When the model exsts heteroscedastcty or seral correlaton, the estmator s stll unbased and consstent, but not effectveness under same varance. Therefore, t s necessary to re-estmate the model

8 by usng feasble generalzed least squares (FGLS) of a relax assumpton. In table 2, all parameters are statstcal sgnfcant wthn the level of 10%. Wald test also makes the same vew clear. Table 2 Estmaton results of FGLS exsted n heteroscedastcty and frst-order autocorrelaton Independent varable Coeffcent Standard error lnt *** lnl * lnk ** lnw ** Constant *** Varable number 310 Wald test χ 2 = Prob>χ 2 =0.000 Note: 1 Agrcultural output value s seen as dependent varable 2 *ndcates a 10% sgnfcant level, ** ndcates a 5% sgnfcant level and *** ndcates a 1%. Calculaton Results For average change rate of varables, we use the followng formula: a(1 + n) t = b (10) Here, a s ntal (1997) varable value; b s termnal (2006) varable value; t, the perod of growth, s 9 and n represents average growth rate. Accordng to the above conclusons, α s and β s Then through the formula (9), we can obtan the "growth drag"( ) of water resources to Chna's agrculture. In other words, owng to the shortage of water resource, Chna's agrculture s greatly affected and productvty s droppng. In addton, water resource can t be up at the same proporton wth crop areas, whch makes the growth rate of agrcultural output reduced % per year on average. Ths also proves that Chna's agrcultural water consumpton s of depleton. In the long run, the ssue should be closely concerned. The prevous quanttatve analyss show, due to nsuffcent supply of water resource, the average output growth rate of Chna s agrculture s contnuously decreasng from 1997 to The numercal value does not seem large, but, accordng to the projecton, by 2030, Chna s agrcultural output growth rate per acre wll lower 2.66% than the current because of the "growth drag", lower 3.74% by 2040 and 4.82 % by Therefore, water resource wll play an mportant role on Chna s agrcultural producton, and wll cause greater nfluence to agrcultural sustanable development. CONCLUSIONS Water, as a renewable resource, exsts growth drag n agrcultural producton. Our fndngs show

9 the growth drag has slowed down the output growth rate of lands, lmtng the gran yeld. In addton, ferce competton from ndustry and cty, together wth clmate warmng and water polluton, also makes useable water resource be greatly reduced, whch wll further ntensfy the harm degree of growth drag. Thus, from a long-term perspectve, f t can not be effectvely controlled, gradually accumulated, t s bound to cause serous damage to Chna's agrcultural producton. In addton, from the formula (9), we can see the "growth drag" s postvely related wth output elastcty α of captal, output elastcty β of water resource and grow rate of crop areas, whch ndcates the "drag" can be reduced through α and β. Its ntrnsc economy meanng s that we can acheve our am through lowerng the role of captal, land and water resource of agrcultural producton. That s to say, agrcultural development can t overly depend on captal, land and water resource. But t does not mean wastng water resources randomly. The appearance of "growth drag" s just because water resource s relatvely shortage and can t grow at the same proporton wth land. Therefore, we must try to reform the current use patterns of agrcultural water, rely on technologcal progress, protect and use water resource ratonally, adopt advanced water-savng technology and measures to mprove water-usng effcency. Only by dong so can we allevate ncreasngly serous water-usng pressure and mantan sustanable growth of natonal economy wthout ncreasng or even decreasng gross water. References Ma Xaohe, Fang Songha.(2006).Chna's water resources and agrcultural producton. Chna's rural economy, (10): Brown L R, Halwel B(1998). Chna's water shortage could shake world food securty. World Watch, 11(4): Davd Romer(2003). Advanced Macroeconomcs. Wang Genpe translaton. Second edton (Chnese translaton). Shangha: Shangha Unversty of Fnance and Economcs Press, Wang Zheng, Xe Shulng,Zhu Janwu(2004). The "growth tal effect" analyss of Chna's economc development. Fnancal Research, 30(9): Xe Shulng, Wang Zheng, Xue Junbo(2005). The "growth tal effect" analyss of land and water resource durng Chna's economc development. Management world, (7): Yang Yang, Wu Cfang, Luo Ganghu(2007). The research of Chnese water and sol resource to economy growth dampng. Economc geography, (4):

10 Cu Yun(2007). The tal effect analyss of land resource n the process of Chnese economc growth. Economc theory and economc management, (11): Kanekos, Tanakak, Toyotat(2004). Water Effcency of Agrcultural Producton n Chna: Regonal Comparson from 1999 to Internatonal Journal of Agrcultural Resources, Governance and Ecology, (3): Zhuo Haohu( ). Carry out unswervngly the constructon of farmland and water conservancy.