216 Internatonal Conference on Mathematcal, Computatonal and Statstcal Scences and Engneerng (MCSSE 216) ISBN: 978-1-6595-96- he Effects of Industral Structure Change on Economc Growth n Chna Based on LMDI Decomposton Approach Yu-yng JIANG *, Er-yan ZHANG and L-yun PAN Department of Basc Scence, Bejng Insttute of Graphc Communcaton, Bejng, Chna *Correspondng author Keywords: Industral structure, Industry growth effect, Structural change effect. Abstract. In ths paper, the total effect of economc growth changes s decomposed nto three ndustry growth effect and structural change effect through LMDI decomposton method n order to probe Chnese ndustral structure change snce the 28 fnancal crss on the contrbuton of economc growth. he results show that the contrbuton share and proporton of the tertary ndustry growth are postvely to the correlaton, and the proportons of the prmary and secondary ndustry growth are negatve correlaton. Introducton Snce the reform and openng up, Chna's economc growth has made outstandng achevements all over the world. However, snce the global fnancal crss n 28, the economc growth has such problems as nsuffcent nternal mpetus drvng, acute structural mbalance. A great development has taken place n the ndustral structure under the market mechansm and government s gudance. From 28 to 214, the proporton of the value-added of prmary ndustry n our economc aggregate s reduced to 9.2% from 1.%, and secondary ndustry s reduced to 46.8% from 42.6%, whle the added value of the tertary ndustry proporton s ncreased to 42.9% from 42.9%.he effects of ndustral structure change to our country economc growth, many scholars has carred on the emprcal research, such as J (26), Sun (211), Gan (211), Deng (21), etc. he study on upgradng of an ndustral structure held a postve atttude to promote Chna's economc growth, whch helps us to understand the structure of dvdend "role n economc growth. However, the research s generally adopts the models of multnatonal or Shft - Share Method for analyss. he selecton of control varable n the multnatonal model has a certan ambguty. he shft -Share method s an effectve method of analyss of structural problems, but t wll underestmate the strength of the structure effect. In ths paper, we proposed a decomposton analyss method, based on Logarthmc Mean Dvsa Index Method (LMDI), from the dynamc change of GDP growth to buld decomposton model of GDP growth. he total effect of economc growth changes s decomposed nto two parts of ndustry growth effect and structural change effect, n order to analyze the mpact of ndustral structure change on Chna's economc growth. Modelng and Data Explanaton he basc economc theory shows that a country's natonal economy can be decomposed nto the sum of the three ndustry, then Y ( GDP t ) Y ( ) ( ) ( ) 1 t Y2 t Y t, (1) where YGDP() t, Y ( t)( 1,2,) represent ncrements of GDP and prmary, secondary and tertary ndustres ncrements n the t th tme respectvely. On both sdes of the formula (1), we dvde Y GDP ( t 1), the value of GDP, then we can get the economc growth decomposton formula as follows: 45
Y ( t) Y Y ( t 1) R ( t) R ( t) f ( t) GDP t, GDP YGDP ( t 1) 1 Y ( t 1) YGDP ( t 1) 1 where ( 1), (2) Yt s the th ndustres ncrements value n the (t-1) th perod, Rt () s the growth rate of the th ndustral sector n the t th perod, f() t s the proporton of the th ndustral sector n total economy at the begnnng of the t th perod (.e., he end of (t-1) th perod ). In order to decompose economc growth rate over tme, we fnd the dervatve wth respect to tme n formula (2), RGDP ( t) R ( t) f ( t) f( t) R( t). () t 1 t 1 t hus, changes n economc growth can be decomposed nto two effects nfluence. On the rght sde of the formula (), the frst term s the ndustry growth effect (.e., Changes n overall economc growth caused by changes n the growth rate of all ndustres f other condtons reman unchanged); the second term s the structural change effect (.e., Changes n economc growth rates caused by changes n the proporton of each ndustry n the natonal economy f other condtons reman unchanged). Integratng on both sde of formula (), we can get: R ( ) t f( t) ( RGDP ), f ( ) ( ) t dt R t dt t t 1 1 (ln R ( t)) (ln f ( t)) R ( t) f ( t) dt R ( t) f ( t) dt 1 t 1 t (ln R ( t)) (ln f( t)) ( ) w( t) dt t w t dt 1 t 1 ( R ) gro str GDP, R ( GDP ), where, ( R ) R ( ) R (), (5) GDP, GDP GDP w ( t) R ( t) f ( t). (6) In formula (4), wt () represent the weght of the th ndustral sector n the overall economc growth, and ts value changes at each tme pont of tme [, ]. hus, we cannot get the value of the effect of ndustral growth and ndustral structure changes through the calculaton on the rght sde of formula (4). Let w ( t) w, then (ln R ( t)) R ( ) ( R ) w dt w ln( ), (7) () gro * * GDP, 1 t 1 R (ln f ( t)) f ( ) ( R ) w dt w ln( ). (8) () str * * GDP, 1 t 1 f he form of the decomposton model depends on the determnaton method of w. Here, referrng to the method of determnng the weght value of the LMDI decomposton method proposed by Ang(21), let w be the log-mean value of the end pont of the nterval,.e.: * * w( ) w() w. (9) w ( ) ln( ) w () Use formula (9) to determne the value of ( R GDP ),. (4) * w, we can acheve a complete decomposton of 451
Economc statstcal data ndcate that after the 28 fnancal crss, Chna's economc growth started to slow down, especally snce the frst quarter of 21, Chna's quarterly GDP growth rate showed a gradual declne n trend, decreased by 11.9% n the early to 6.9% n the thrd quarter of 215. Chnese economy has entered a new normal n the hgh-speed growth. he slowdown n Chna's economc growth s a structural slow down, the new normal of Chnese economc growth should be the growth of an optmzed and upgraded ndustral structure. herefore, ths paper chooses the perod of Chna's economc slowdown after the fnancal crss for the study perod. Usng the decomposton analyss method to nvestgate the mpact mechansm of ndustral structure change on economc growth, we can provde reference for the future of ndustral structure optmzaton under the Chna's new normal growth. We select the natonal quarter GDP level and the proporton of the growth value of the three ndustry of the frst quarter of 27 to the thrd quarter of 215 as the orgnal data. akng the comparablty of the data nto account, quarterly GDP data s used the 21 constant prce measure. Snce the Natonal Bureau of statstcs only released the GDP constant prce quarterly data after 211, therefore, for the constant prce level of GDP before 211, we can calculate them wth the help of the year-on-year growth rate of total amount calculated at constant prces GDP publshed at the same perod, then calculated the ncrease of values at that perod accordng to the proporton of three ndustres over the same perod for each ndustry at constant prce measure. Because of the decomposton of the model s based on the economc growth rate, so we further calculate the year-on-year growth rate wth the help of constant prce GDP and quarterly ncreased value of three ndustres at constant prce measure (the same quarter n last year as the base perod). hen, we can effectvely avod the nfluence of seasonal factors n the model decomposton calculaton. Emprcal Analyss Usng formula (5), (7) and (8) to calculate the change of the quarterly GDP year-on-year growth rate, we can get the total effect of economc growth changes ( R GDP ),, ndustry growth effect ( R ) gro GDP, and structural change effect ( R ) str GDP,. he th perod of each decomposton calculaton s the same quarter of last year. he decomposton results are shown n table 1. otal Effect he observaton of the change of total effect can help us to have a whole understandng of the trend of economc growth after the fnancal crss. In table 1, the total effect calculaton results show that, after the fnancal crss, economc growth rate showed an overall downward trend, the average contrbuton of each perod of total effect s -.18 percentage ponts, but from the path of change, the economc growth frst experenced a short-term adjustment and shock, after whch began to enter a stable declne phase. By the drect mpact of the fnancal crss, from the frst quarter of 29 to the thrd quarter, economc growth has a rapd declne perod, quarter year-on-year GDP growth average declne of 2. percentage ponts. And wthn the next four quarters, economc growth has a short-term rebound, quarter year-on-year GDP growth ncreased by an average of.4 percentage ponts, reachng the level of growth before the fnancal crss, but then (from the fourth quarter of 21 to the thrd quarter of 215) economc growth started to slow down, the level of total effect almost all be negatve, quarterly year-on-year GDP growth s reduced by an average of.76 percentage ponts, Chna's economc growth went nto the new normal perod. 452
able 1. Decomposton results of the effects of ndustral growth and structural change on economc growth. Change nterval otal effect Industral growth effect Structural change effect [, ] ( ), R GDP ( R ) gro GDP, ( R ) str GDP, 8,1~9,1 -.45 -.448 -.19 8,2~9,2 -.2246 -.2219 -.26 8,~9, -.257 -.255 -.1 8,4~9,4.44.4415.18 9,1~1,1.58.576.94 9,2~1,2.24.27.9 9,~1,.959.864.95 9,4~1,4 -.1659 -.1681.2 1,1~11,1 -.22 -.224.4 1,2~11,2 -.897 -.915.19 1,~11, -.68 -.681.1 1,4~11,4 -.86 -.86. 11,1~12,1 -.174 -.17 -.1 11,2~12,2 -.1991 -.199.2 11,~12, -.161 -.167 -.4 11,4~12,4 -.67 -.69.17 12,1~1,1 -.229 -.257.28 12,2~1,2 -.62 -.19.77 12,~1,.47.7.1 12,4~1,4 -.84 -.465.8 1,1~14,1 -.48 -.581.11 1,2~14,2 -.14 -.199.95 1,~14, -.677 -.752.75 1,4~14,4 -.99 -.459.6 14,1~15,1 -.275 -.64.89 14,2~15,2 -.46 -.478.72 14,~15, -.4 -.425.82 average -.18 -.62.45 Industral Growth Effect From the overall decomposton results n able 1, the ndustral growth effect s stronger than the structural change effect, whch plays a decsve role n the economc growth. he average contrbuton of ndustry growth n the whole research area s percentage ponts, and the ndustry growth rate has a sgnfcant negatve effect on the overall economc growth. Structural Change Effect As shown n able 1, the average contrbuton of ndustral structure change effect n the study area s.45 percentage ponts, whch ndcates that the change of ndustral structure has postve effect on economc growth, but the effect s not very sgnfcant. In partcular, there exst fve phase whch structure change effect s negatve before 212, shows that the adjustment of ndustral structure s not 45
reasonable n ths perod. After that, the structure changes effect s always postve and more stable. It shows that the ndustral structure adjustment of our country makes the allocaton of resources n the ndustry steadly move n a reasonable drecton, and so releve the overall economc downturn trend. Conclusons hrough the decomposton analyss of Chna's economc growth after the fnancal crss, we draw the followng man conclusons: Frst, Chna's ndustral structure has begun to evolve stably to ratonalzaton, and has played a postve role n the natonal economc growth n the context of the overall slowdown n economc growth. However, the pullng effect s much smaller than the downward pressure caused by the declne of three ndustral growth rate, can only play a very small role n easng the overall economc slowdown. Secondly, after the fnancal crss, the declne n three ndustral growth rate has become a major factor n hnderng economc growth. o mantan the sustaned growth of the natonal economy, we stll need to accelerate the structural adjustment and upgradng of the ndustral structure, lookng for the drvng force of each ndustry growth, n order to mantan the stablty of the ndustral growth. Fnally, among the change n the three ndustral structure, changes of proporton of the thrd ndustry s the most strongly correlated to the natonal economc growth. In the future, we would contnue to ncrease the proporton of the thrd ndustry to make t as a new man drvng force of economc growth. Acknowledgement hs research was fnancally supported by BIGC Project (Eb2166, Ea2166). References [1] Deng B. he nfluences of ndustry structure changes to Economc growth. J. Xangtan Unv.Phl.Soc.Sce., 21, 7(5):4-47.(In Chnese) [2] Gan C.H. et al. An Emprcal Study on the Effects of Industral Structure on Economc Growth and Fluctuatons n Chna, Economc Research Journal, 211(5):4-16. (In Chnese) [] J Y.,Wu Y. On the contegraton-model of the relatons between ndustral structure and economc growth n Chna. Contemporary Economc Research, 26(6):47-51. (In Chnese) [4] Sun H., Sh Zh.X. he Industral Structure and Economc Growth of Chna: A Research Based on the Industry Labor Rate. Populaton and Economcs, 211(2):1-6. (In Chnese) [5] Wang Q. et al. Emprcal Research on the Drven Relatonshp Between R&D Input and Economc Growth of Our Country: Based on LMDI Decomposton Method. Economc survey, 211(5):18-21. (In Chnese) [6] Ang. A new energy decomposton method: perfect n decomposton and consstent n aggregaton. Energy, 21, 26(6):57-548. [7] Chenery. Patterns of Industral Growth. Economc Growth n the Long Run A Hstory of Emprcal Evdence, 196, 28():69-691. [8] Chenery. A Comparatve Analyss of Industral Growth[M]. Palgrave Macmllan UK, 198. [9] Kuznets. Quanttatve aspects of the economc growth of natons: ndustral dstrbuton of natonal product and labor force. Economc Development and Cultural Change.1957, 5(4):1-111. [1] Peneder. Industral structure and aggregate growth. Structural Change and Economc Dynamcs, 2, 14(4):427-448. 454