Emprcal Sudy on he Relaonshp beween ICT Applcaon and Chna Agrculure Economc Growh Pengju He, Shhong Lu, Huoguo Zheng, and Yunpeng Cu Key Laboraory of Dgal Agrculural Early-warnng Technology Mnsry of Agrculure, People s Republc of Chna, Bejng 100081, P.R. Chna Absrac. In recen years, Chnese governmen aaches grea mporance o nformaon and communcaon echnology ICT applcaon n agrculure. Drven by marke profs and fnancal prvlege, nformaon and communcaon enerprses, agrculural enerprses, research nsues, unverses and relave assocaons have acvely nvolved n ICT applcaon n agrculure for many years. Therefore, sudy on relaonshp beween ICT applcaon and Chna agrculure economc growh s of sgnfcance. Ths paper esablshed a new model on agrculure economc growh and usng panel daa made an emprcal sudy on he relaonshp beween ICT applcaon and chna s agrculure economc growh. Keywords: nformaon and communcaon echnology applcaon; agrculure; economc growh; panel daa; regresson. 1 Inroducon Alhough here are few research resuls on he relaonshp beween ICT applcaon and agrculure economc growh, scholars have acheved some mporan resuls on he relaonshp beween ICT applcaon and economc growh, he mehodology of whch s also applcable. Sudes on he relaonshp beween ICT applcaon and economc growh followed wo drecons. The frs drecon s from he perspecve of ndusral economcs. Represenave sudes nclude: Marc U, Pora (1997) dvdes nformaon secor no prmary nformaon secor and secondary nformaon secor, usng npu-oupu able mehod, he measures he sze of he U.S. nformaon economy [1]. The second drecon s from he perspecve of producon funcon, whch deems ICT applcaon as an economc npu facor. Represenave sudes nclude: Gurmukh Gll, ec 1997 collecs daa n 11 cross-ndusry areas from 1983 o 1993 of 58 ndusres of Uned Saes and uses producon funcon analyzng he conrbuon of nformaon echnology [2]. Inernaonal Telecommuncaon Unon n 2006 publshed "World Telecommuncaon/ICT Developmen Repor" (2006) whch demonsraes he nformaon and communcaon echnologes can grealy mprove producvy and made an concluson ha he rue poenal of nformaon and communcaon echnology s no s drec mpac on he economy, bu he ndrec mpac on he whole economc sysem [3]. Welfens (2002) decomposed scenfc and echnologcal D. L, Y. Lu, and Y. Chen (Eds.): CCTA 2010, Par III, IFIP AICT 346, pp. 648 655, 2011. IFIP Inernaonal Federaon for Informaon Processng 2011
Relaonshp beween ICT Applcaon and Chna Agrculure Economc Growh 649 advances no nformaon echnology advances and non-nformaon echnology advances [4]. Formula 1 shows he relaonshp beween echnologcal progress, nformaon echnologcal progress and non-nformaon echnologcal progress: A =A 0 Z γ (1) In formula 1, Z sands for he level of nformaon echnology advances. γ sands for he oupu elascy of nformaon echnology advances. A 0 sands for he advances of non-nformaon echnology. In chna, researches on he relaonshp beween ICT applcaon and economc growh are also followed he wo drecons, noeworhy sudes ncludes: Zhu Youpng (1996) uses C-D producon, n whch he dependen varable s Chna's real GDP and he ndependen varables are physcal capal, labor force and nformazaon facor, collecng daa from 1980 o 1992 o do regresson analyss and fnds ou ha he nformazaon facor makes he larges conrbuon of chna s economc growh [5]. Tao Changq (2001) does correlaon analyss among nformaon echnology and equpmen manufacurng, nformaon servces and radonal ndusry and fnds ou ha correlaon coeffcen s of sgnfcan [6]. Researches on he relaonshp beween ICT applcaon and agrculure economc growh are comparavely few. Noeworhy sudes nclude: Zhang Hong, Zhang Quan (2006) frsly uses he mehod of "nformaon echnology ndex n Japan" o measure Chna's rural nformaon and hen consruc a model whch seems rural nformazaon as a new facor. Furher more, hey conduc an emprcal sudy on Chnese agrculural developmen from 1993 o 2002 and fnd ou rural nformazaon plays an mporan role n chna s agrculural developmen [7]. Alhough hese sudes menoned above are useful, we can fnd ou ha here are sll some lmaons: frsly, mos ndex sysem for measuremen of ICT s no suable n chna, because some mporan ndcaors are no covered n chna s sascs. Secondly, here are huge dfferences of ICT applcaon level n dfferen regons of chna, mos sudes akng chna as a whole whou hypohess esng o fnd ou he relaonshp beween ICT applcaon and agrculure economc growh s naccurae. Ths paper measures ICT applcaon level n a dfferen way, n whch consumpon and nvesmen on ICT applcaon s used as ndcaors of ICT applcaon level. Then, hs paper collecs daa n 30 provnces of chna and usng panel daa analyzes he relaonshp beween ICT applcaon and agrculure economc growh. Ths paper s organzed as follows: Frsly, hs paper makes a leraure revew on he relaonshp beween ICT applcaon and economc growh and pu forward he research oulne of hs paper. Secondly, based on sysem engneerng and nformaon economc analyss, hs paper esablshed a new model, whch sees ICT applcaon as a new facor, on agrculure economc growh and makes an nerpreaon abou he varables. Thrdly, usng panel daa from 1999 o 2006 n 30 provnces, hs paper made an emprcal sudy on he relaonshp beween ICT applcaon and chna s agrculure economc growh. A las, hs paper pus forward some suggesons on ICT applcaon n chna s agrculure.
650 P. He e al. 2 Model and Varables Selecon 2.1 Model Consrucon Based on he research resuls menoned before, he agrculure economc growh model of hs paper s bul on hree hypoheses. Frs, ICT applcaon s a new ndependen facor of agrculural economc sysem; second, ICT applcaon has a posve exernaly on agrculure economc growh; hrd, ICT applcaon facor s an endogenous varable of agrculural economc sysem. Based on hese hree hypoheses, he form of agrculure producon funcon can be expressed as follows: Y=F A 0, K, H, I (2) In he model, Y sands for he oal agrculure economy oupu, A 0 sands for scenfc and echnologcal progress removng elemens of he agrculural and rural nformaon, K sands for physcal capal nvesmen, H sands for human capal nvesmen, I sands for ICT applcaon. For he specfc form of producon funcon, hs sudy uses Cobb Douglas form. Pu he me varable no he model, he model used n hs sudy s showed as follows: Y A 0 α β γ = K H I (3) Afer necessary mahemacal reamens of model 3, model 4 deduced from Model 3 s he sascal model o use: Y ' = C + K ' + β H ' + γ I ' + α μ y + ξ (4) 2.2 Varables Selecon and Daa Processon Dependen varable Y: In hs sudy, he oal agrculural producon was used o measure he oupu of he agrculural economy. The daa can be colleced from he em "prmary ndusry of gross domesc produc" (GDP) of "Chna Sascal Yearbook". In order o elmnae he mpac of prce facor, wh 1990 prces as base year, "he frs ndusral GDP ndex" of "Chna Sascal Yearbook" was used o adjus he nomnal value of prmary ndusry of gross domesc produc no acual values. Physcal capal K: In hs sudy, he sock comng from fxed asses nvesmens on he prmary ndusry was used as a measure of agrculural physcal capal. As here s no drec sascal daa on agrculural physcal capal sock n he counry, hs sudy as a common way used he perpeual nvenory mehod o calculae, he formula s: K = 1-δ K 1 +I. In he formula, K and I sand for he capal sock and new nvesmen n year respecvely, δ sands for he deprecaon rae. Accordng o some research resuls, δ s assumed o be 5%. For agrculural nvesmen, I can be colleced from he fxed asse nvesmen daa on agrculure, foresry, anmal husbandry and fshery of "Chna Rural Sascal Yearbook". Ths sudy uses he mehods developed by Hall and Jones (1999) and Young (2000) o calculae he base perod capal sock,
Relaonshp beween ICT Applcaon and Chna Agrculure Economc Growh 651 he formula s: K 0 =I 0 /(g + δ), n whch, I 0 sands for he nvesmen of he base perod, g sands for he annual average growh rae for nvesmen. In order o elmnae he mpac of prce facor, hs sudy, wh he prces of 1990 he base year, used he nvesmen ndex of each regon o adjus he nomnal value no acual value. Human capal H: As hs sudy uses he human capal ndcaors, raher han he labor force, hus needs o ranslae dfferen ypes of labor force no human capal sock. The mehod of s gve dfferen human capal equvalen coeffcens o people of dfferen educaonal levels. The human capal equvalen coeffcens used n hs sudy are comng from research resuls by Zhou Xao, Zhu Nong (2003) [8]. Human Capal equvalen coeffcens of all levels of educaon are shown n Table 1. Table 1. Human Capal Coeffcens Illeracy Prmary school Junor school Hgh school Secondary school College school and above 1 1.07 1.254 1.1308 1.45* 1.624 ICT applcaon facor I: ICT applcaon facor can be measured by he sum of consumpon and nvesmen on ICT applcaon, whch are more drec and accurae. The nvesmen daa on ICT applcaon comes from hree ems, whch are nvesmens on Informaon ransmsson, compuer servces and sofware, Educaon and a par of Culure, Spors and Eneranmen n Chna Rural Sascal Yearbook. The consumpon daa on ICT applcaon comes from wo ems, whch are nvesmens on communcaon expenses from Transpor and communcaon expenses and Culural and eneranmen producs and servces expendure. In order o deermne he percenage of he nvesmen on Culure, Spors and Eneranmen and he consumpon on Transpor and communcaon expenses, hs sudy conduced surveys n 30 provnces and made esmaes on he paral. In order o elmnae he prce facor, he daa of nvesmen on ICT applcaon and he daa of consumpon on ICT applcaon were adjused by nvesmen ndex and consumpon ndex. As he same, hs sudy adoped he perpeual nvenory mehod o calculae he ICT fxed asse sock n each provnce. 3 Emprcal Sudy and Resuls In order o analyze he mpac of ICT applcaon on agrculure economc growh, hs paper made a comparave sudy. Model 5, whch s a classcal economc growh model, does no nclude ICT applcaon facor, whle model 6, whch s consruced n par 2, ncludes ICT applcaon facor. Model 5 and Model 6 are shown as follows: Y ' C + α K ' + β H ' + μ y + ξ (5) =
652 P. He e al. Y ' = C + K ' + β H ' + γ I ' + α μ y + ξ For model 5, he dependen varable s he real GDP of agrculure deparmen. Through correlaon analyss, can be found ha hese hree ndependen varables ncludng he sock of physcal capal, human capal sock, ICT applcaon facor and regonal dummy varable exss srong auocorrelaon, herefore, he regonal dummy varable was excluded. Model 6, wh he same reason, also excluded he regonal dummy varable. Because he mpac of ICT applcaon lags behnd he nvesmen and consumpon, so ICT applcaon varable daa lag a year n boh model 5 and model 6. Usng he daa n 30 provnces of Chna form 1999 o 2006, hs paper conducs an emprcal sudy on Chna's agrculural economc growh. Through he F es and Hausman es, boh models should use me fxed effecs, common cross-secon coeffcens model o do regresson analyss. I should be noe ha cross-secon weghng was used boh n model 5 and model 6. In hs paper, he sofware whch s used s Evews5.0. Regresson resuls of model 5 and Model 6 were shown n Table 2, 3. In model 5, he R-square, Adjused R-square are 0.999746, 0.999736 respecvely. F-sasc value s 100678.4 and he Probably of F-sasc value s 0.000000. Durbn-Wason sasc value s 0.54. The elasc coeffcens of physcal capal, human capal elemen are 0.49,0.46 and he sum of hem s 0.95 approxmaely equal o 1, ndcang ha for each addonal un of regonal physcal capal, human capal npu can smulae agrculural growh 0.49, 0.46 uns, whch bascally conssen wh he assumpon ha consan reurns o scale. Reference o relevan leraure, he regresson resuls s conssen wh oher sudes: he elasc coeffcen of physcal capal s lager han human capal. However, he resuls show a new feaure ha he elascy of human capal s relavely hgh, ndcang ha human capal plays a growng role n he agrculural economy growh a hs perod. (6) Table 2. Model 5 Regresson Analyss Resuls Varable Coeffcen Sd. Error -Sasc Prob. C 6.327534 0.142652 44.35655 0.0000 K1? 0.491837 0.014703 26.65073 0.0000 H1? 0.459361 0.015842 28.99651 0.0000 Fxed Effecs (Perod) 1999--C 0.055869 2000--C -0.066871 2001--C 0.176475 2002--C -0.055995 2003--C -0.069044 2004--C 0.035609 2005--C -0.013762 2006--C -0.053886
Relaonshp beween ICT Applcaon and Chna Agrculure Economc Growh 653 Table 3. Model 6 Regresson Analyss Resuls Varable Coeffcen Sd. Error -Sasc Prob. C 5.693355 0.152987 37.21457 0.0000 K1? 0.452439 0.017724 19.88461 0.0000 H1? 0.533188 0.023175 23.00680 0.0000 ICT?(-1) 0.154849 0.023945 6.466894 0.0000 Fxed Effecs (Perod) 2000--C -0.062033 2001--C 0.175993 2002--C -0.043696 2003--C -0.048470 2004--C 0.059661 2005--C 0.016271 2006--C -0.046159 In model 6, he R-square, Adjused R-square are 0.999871, 0.999866 respecvely. F-sasc value s 172870.6 and he Probably of F-sasc value s 0.000000. Durbn-Wason sasc value s 0.62. The regresson resul s neresng. Frs, ICT applcaon facor pas hrough he 1% sgnfcance hypohess esng, whch means ha ICT applcaon has become one of he elemens of agrculural economc growh. Second, compared o model 5, he elascy of physcal capal s almos unchanged, whle he elascy of human capal s ncreased o be larger han he elascy of physcal capal and he sum of hem, whch s 0.99, sll approxmaely equal o 1. Thrd, he elasc coeffcens of physcal capal, human capal, ICT applcaon facor are 0.45, 0.53, 0.15 and he sum of hem s 1.13 larger han 1, ndcang ha for each addonal un of regonal physcal capal, human capal, ICT applcaon facor npu can smulae agrculural growh 0.45, 0.53, 0.15, whch means some ncreasng reurns o scale and ICT applcaon facor ncreases reurns o scale. Comparng Model 6 and Model 5, on he one hand, we can see ha here s a srong conssency of he wo models. Frs, he regresson resuls of he wo models are consan n esmaes of he elascy of physcal capal and human capal, whch he dfferences beween hem are comparavely small and he sum of hem approxmaely equal o 1. Second, he regresson resuls of he wo models are consan n esmaes of me fxed effecs. The meanng of me fxed effecs s clearly, apar from he facors ncludng scenfc and echnology progress, physcal capal npu and human capal npu, should be naural condons and he effecs of he polces. Accordng o model 6, he me fxed effecs from 2000 o 2006 were -0.062033, 0.175993, -0.043696, -0.048470, 0.059661, 0.016271, -0.046159, whch are reasonable and can be explaned by naural condons and polces. Accordng o he acual agrculural producon, 2001 s a favorable weaher year n Chna and hs year's agrculural producon made a grea growh. 2005 s a grea me for Chnese farmer, n March 2005, here had been 26 provnces (auonomous regons and muncpales) announced he abolon of agrculural ax, whch smulaes agrculure oupu growh. On he oher hand, here
654 P. He e al. are dfferences. Frs, we know ha he consan em n model 5 sands for scenfc and echnologcal progress, whle he consan em n model 6 sands for scenfc and echnologcal progress excludng ICT applcaon facor. From he resuls, can be seen ha he consan em n model 5 s lager han he consan em n model 6, whch s reasonable. Second, he ICT applcaon facor has changed he npu-oupu srucure of he agrculural economy. The ICT applcaon facor has become one ndependen elemen n he agrculural economc sysem and ncreases he conrbuon of human capal nvesmen o agrculure economc growh. By comparng he resuls, can be concluded ha boh model are reasonable, bu model 6 holds more ruh. 4 Dscussons Through he analyss of agrculural economy growh, we can draw hree mporan conclusons: Frs, ICT applcaon has become an mporan facor n agrculure economc growh. Second, ICT applcaon has changed he oupu elascy of physcal capal and human capal, whch renforces he role of human capal, herefore changed he way of economc growh n agrculure. Thrd, ICT applcaon has mproved he agrculural economc growh rae reurns o scale and has become he source of economes scale ncremen of agrculure. Therefore, In order o, accelerae he developmen of agrculural economy, he decson-makers should no only o consder he physcal capal nvesmen, human capal nvesmen and echnologcal progress n agrculure felds, bu aach grea mporance o promoe agrculural and rural nformazaon developmen. Meanwhle, can be seen ha he mpac of ICT applcaon on physcal capal nvesmen s no obvous. Therefore, n he agrculural nformazaon consrucon process, we should pay more aenon o develop nellgen equpmen, whch can drecly mprove he effcency of agrculural producon, and agrculural nformazaon consrucon. There s a need o encourage scenfc research nsuons, nformaon echnology, companes and agrculural enerprses o make effors o mprove he nellgence level of Chnese agrculural producon ools and facles, herefore promoe agrculural growh. References 1. Pora, M.U.: The Informaon Economy (9 volumes), Offce of Telecommuncaons Specal Publcaon 77-12, Washngon D.C.: U.S. Deparmen of Commerce (1977) 2. Gll, G., Young, K., Pasore, D., Dumagan, J.C., Turk, I.: Economy - Wde and Indusry - Level Impac of Informaon Technology (4) (1997) 3. ITU. World elecommuncaon developmen repor 2006: Measurng ICT for Socal and Economc Developmen (2006) 4. Welfens, P.J.J.: Informaon & Communcaon Technology and capal marke perspecves. In: Inernaonal Economcs and Economc Polcy, vol. 2(1), Sprnger, Hedelberg (2005) 5. Zhu, Y.: Informaonalzaon and Is Influence on economy growh. Journal of he Chna Socey for Scenfc and Techncal Informaon (5) (1996)
Relaonshp beween ICT Applcaon and Chna Agrculure Economc Growh 655 6. Tao, C.: The Real Analyss of Informaon Indusry of Our Counry. Journal of Eas Chna JaoTong Unversy (04) (2001) 7. Zhang, H., Zhang, Q.: Rural nformazaon mpac on Agrculural economc growh. Sascs and Decson (12) (2008) 8. Zhou, X., Zhu, N.: Reurn o Human Capal n Rural Chna. Chnese Journal of Populaon Scence (06) (2003)