謝雄軍++何紅渠??
摘 要:利用1985~2011年我國(guó)各個(gè)省域產(chǎn)業(yè)集聚和經(jīng)濟(jì)發(fā)展面板數(shù)據(jù)進(jìn)行了空間計(jì)量分析,依次包括全域空間自相關(guān)分析、局部空間自相關(guān)分析、空間面板滯后模型分析、空間面板誤差模型分析等。實(shí)證分析結(jié)果表明:我國(guó)產(chǎn)業(yè)集聚和區(qū)域經(jīng)濟(jì)增長(zhǎng)具有明顯的區(qū)域差異;產(chǎn)業(yè)集聚與區(qū)域經(jīng)濟(jì)增長(zhǎng)均存在比較明顯的空間自相關(guān);產(chǎn)業(yè)集聚對(duì)區(qū)域經(jīng)濟(jì)增長(zhǎng)的影響具有正向空間自相關(guān)特性,即產(chǎn)業(yè)集聚對(duì)區(qū)域經(jīng)濟(jì)增長(zhǎng)具有顯著促進(jìn)作用。
關(guān)鍵詞: 產(chǎn)業(yè)集聚; 省域經(jīng)濟(jì);空間計(jì)量;空間面板數(shù)據(jù);Morans I
中圖分類號(hào):F061 文獻(xiàn)標(biāo)識(shí)碼: A 文章編號(hào):1003-7217(2014)02-0116-06
一、引 言
產(chǎn)業(yè)集聚是指同一產(chǎn)業(yè)在某個(gè)特定地理區(qū)域內(nèi)高度集中,產(chǎn)業(yè)資本要素在空間范圍內(nèi)不斷匯聚的一個(gè)過(guò)程。產(chǎn)業(yè)集聚問(wèn)題的研究產(chǎn)生于l9世紀(jì)末,馬歇爾在1890年就開始關(guān)注產(chǎn)業(yè)集聚這一經(jīng)濟(jì)現(xiàn)象,并提出了兩個(gè)重要的概念即“內(nèi)部經(jīng)濟(jì)”和“外部經(jīng)濟(jì)”。馬歇爾之后,產(chǎn)業(yè)集聚理論有了較大的發(fā)展,出現(xiàn)了許多流派,比較有影響的有:韋伯的區(qū)位集聚論、熊彼特的創(chuàng)新產(chǎn)業(yè)集聚論、E·M·胡佛的產(chǎn)業(yè)集聚最佳規(guī)模論、波特的企業(yè)競(jìng)爭(zhēng)優(yōu)勢(shì)與鉆石模型等。以產(chǎn)業(yè)園或一定區(qū)域范圍為研究對(duì)象的產(chǎn)業(yè)集聚發(fā)展及其對(duì)經(jīng)濟(jì)增長(zhǎng)影響的研究不勝枚舉。Martin指出集聚與經(jīng)濟(jì)增長(zhǎng)是內(nèi)生的相互促進(jìn)的過(guò)程,集聚降低創(chuàng)新成本從而促進(jìn)區(qū)域經(jīng)濟(jì)增長(zhǎng),同時(shí),經(jīng)濟(jì)增長(zhǎng)的進(jìn)一步成長(zhǎng)也會(huì)反向促進(jìn)新的集聚[1];Brulhart認(rèn)為集聚在早期會(huì)促進(jìn)經(jīng)濟(jì)增長(zhǎng),但發(fā)展到一定階段擁擠效應(yīng)會(huì)造成發(fā)展瓶頸,反向調(diào)節(jié)集聚的水平[2]。國(guó)內(nèi)的韓寶龍也從鄰近性理論角度指出,產(chǎn)業(yè)集聚發(fā)展對(duì)區(qū)域經(jīng)濟(jì)發(fā)展的影響是周期性的自反饋調(diào)節(jié)作用,存在地理鄰近的負(fù)效應(yīng)[3];同時(shí),徐盈之等也通過(guò)集聚對(duì)增長(zhǎng)具有非線性效應(yīng)證實(shí)了“威廉姆森假說(shuō)”[4]。
但是,現(xiàn)有關(guān)于產(chǎn)業(yè)集聚與區(qū)域經(jīng)濟(jì)增長(zhǎng)關(guān)系的探討中都把區(qū)域當(dāng)作獨(dú)立的封閉空間,忽視經(jīng)濟(jì)系統(tǒng)的開放性和多區(qū)域空間之間經(jīng)濟(jì)集聚特征的相互作用。近年來(lái)興起的空間計(jì)量經(jīng)濟(jì)學(xué)方法可以填補(bǔ)這一空白,已有學(xué)者利用空間計(jì)量分析方法研究我國(guó)縣域經(jīng)濟(jì)發(fā)展的空間差異特征[5],以及利用空間滯后模型和空間誤差模型,研究發(fā)現(xiàn)人力資本對(duì)區(qū)域創(chuàng)新有顯著促進(jìn)作用[6]。但尚無(wú)研究將空間計(jì)量分析方法應(yīng)用到我國(guó)產(chǎn)業(yè)集聚空間特征與區(qū)域經(jīng)濟(jì)增長(zhǎng)特征的相關(guān)性分析當(dāng)中,更無(wú)基于面板數(shù)據(jù)的產(chǎn)業(yè)集聚與區(qū)域經(jīng)濟(jì)增長(zhǎng)關(guān)系計(jì)量分析。因此,本文試圖使用面板數(shù)據(jù)對(duì)我國(guó)的產(chǎn)業(yè)集聚特征進(jìn)行時(shí)間和空間的雙重分析,并討論區(qū)域經(jīng)濟(jì)發(fā)展與產(chǎn)業(yè)集聚之間的時(shí)空關(guān)系。
二、我國(guó)產(chǎn)業(yè)集聚及省域經(jīng)濟(jì)發(fā)展的空間特征
(一)產(chǎn)業(yè)集聚的空間分布特征
五、結(jié)論與討論
以上分析表明:(1)我國(guó)產(chǎn)業(yè)集聚和省域經(jīng)濟(jì)增長(zhǎng)具有顯著的區(qū)域差異特征,都表現(xiàn)為由東南沿海向西北內(nèi)陸降低的趨勢(shì)。(2)全域空間自相關(guān)MoranI分析說(shuō)明產(chǎn)業(yè)集聚與省域經(jīng)濟(jì)增長(zhǎng)都存在比較明顯的空間自相關(guān)關(guān)系,同時(shí)對(duì)產(chǎn)業(yè)集聚與省域經(jīng)濟(jì)增長(zhǎng)的雙變量MoranI分析說(shuō)明產(chǎn)業(yè)集聚對(duì)省域經(jīng)濟(jì)增長(zhǎng)的影響具有正自相關(guān)特性。(3)局部自相關(guān)LISA聚類分析發(fā)現(xiàn),各省的產(chǎn)業(yè)集聚空間自相關(guān)分屬于四種聚類類型,北京、天津、廣東、福建等屬于高高類型;山東、江蘇、遼寧等屬于低低類型;陜西、河北、內(nèi)蒙古等屬于高低類型;廣西、湖南、云南等屬于低低類型。省域經(jīng)濟(jì)增長(zhǎng)空間自相關(guān)分屬于四種聚類類型,北京、江蘇、河南等屬于高高類型;貴州、湖南、江西等屬于低低類型;廣東、福建、重慶屬于高低類型,河北和山西等屬于低高類型。(4)對(duì)產(chǎn)業(yè)集聚與省域經(jīng)濟(jì)增長(zhǎng)進(jìn)行空間面板計(jì)量分析,發(fā)現(xiàn)考慮空間自相關(guān)因素的模型在總體模型擬合度和變量系數(shù)顯著性等方面均優(yōu)于不考慮空間自相關(guān)因素的模型,估計(jì)結(jié)果表明在控制了其他影響區(qū)域經(jīng)濟(jì)增長(zhǎng)的因素并考慮到空間自相關(guān)因素后,產(chǎn)業(yè)集聚對(duì)區(qū)域經(jīng)濟(jì)增長(zhǎng)具有顯著促進(jìn)作用。
因此,對(duì)于中西部欠發(fā)達(dá)地區(qū),為實(shí)現(xiàn)向東部發(fā)達(dá)地區(qū)的經(jīng)濟(jì)增長(zhǎng)收斂,可以從擴(kuò)大投資規(guī)模、大力推進(jìn)城市化、加快改革開放步伐、提高產(chǎn)業(yè)集聚水平、注重人力資本投資等方面入手;對(duì)于東部發(fā)達(dá)地區(qū)而言,應(yīng)率先從過(guò)去主要靠投資推動(dòng)的經(jīng)濟(jì)增長(zhǎng)方式向主要依靠人力資本和科技創(chuàng)新推動(dòng)的經(jīng)濟(jì)增長(zhǎng)方式轉(zhuǎn)變。對(duì)產(chǎn)業(yè)集聚和省域經(jīng)濟(jì)水平都處于低低聚類類型省域,需要增強(qiáng)跨省的重大戰(zhàn)略舉措,通過(guò)大區(qū)域發(fā)展來(lái)實(shí)現(xiàn)各省市的經(jīng)濟(jì)實(shí)力;對(duì)于產(chǎn)業(yè)集聚和省域經(jīng)濟(jì)發(fā)展水平空間關(guān)系同處于高低聚類類型或低高聚類類型的省域,需要在產(chǎn)業(yè)鏈建設(shè)上實(shí)現(xiàn)省域間關(guān)聯(lián),通過(guò)產(chǎn)業(yè)鏈合作來(lái)加強(qiáng)鄰近省域的經(jīng)濟(jì)空間關(guān)聯(lián),從而實(shí)現(xiàn)鄰域的耦合發(fā)展。
注釋:
①數(shù)據(jù)來(lái)自于《新中國(guó)六十年統(tǒng)計(jì)資料匯編》、《中國(guó)城市統(tǒng)計(jì)年鑒》(1986-2012)、《我國(guó)人口統(tǒng)計(jì)年鑒》(1986-2012)以及《中國(guó)統(tǒng)計(jì)年鑒》(1986-2012)。
參考文獻(xiàn):
[1]Martin, P., Ottaviano, G.L. Growth and agglomeration [J].International Economic Review,2001,42(4):947-968.
[2]Brulhart, M.,Sbergami, F. Agglomeration and growth:cross country evidence[J]. Journal of Urban Economics, 2012,65(1):48-63.
[3]韓寶龍.地理與認(rèn)知鄰近對(duì)高技術(shù)產(chǎn)業(yè)集群創(chuàng)新影響的實(shí)證研究[D].長(zhǎng)沙:湖南大學(xué)碩士學(xué)位論文,2011.
[4]徐盈之,彭歡歡,劉修巖. 威廉姆森假說(shuō):空間集聚與區(qū)域經(jīng)濟(jì)增長(zhǎng)基于我國(guó)省域數(shù)據(jù)門檻回歸的實(shí)證探討[J].經(jīng)濟(jì)理論與經(jīng)濟(jì)管理,2011,(4):95-102.
[5]吳玉鳴. 縣域經(jīng)濟(jì)增長(zhǎng)集聚與差異:空間計(jì)量經(jīng)濟(jì)實(shí)證探討[J].世界經(jīng)濟(jì)文匯,2007,(2):37-57.
[6]錢曉燁,遲巍,黎波. 人力資本對(duì)我國(guó)區(qū)域創(chuàng)新及經(jīng)濟(jì)增長(zhǎng)的影響基于空間計(jì)量的實(shí)證探討[J].數(shù)量經(jīng)濟(jì)技術(shù)經(jīng)濟(jì)探討,2012,(4):107-121.
[7]Anselin, L., Florax, R. andRey, S. Advanced in spatial econometrics:methodology,tools and applications[M].Berlin:Springer Verlag,2004.
[8]陳建軍,胡晨光.產(chǎn)業(yè)集聚的集聚效應(yīng)以長(zhǎng)江三角洲次區(qū)域?yàn)槔睦碚摵蛯?shí)證探討[J].管理世界,2008,(6):68-83.
[9]王家庭,賈晨蕊. 我國(guó)城市化與區(qū)域經(jīng)濟(jì)增長(zhǎng)差異的空間計(jì)量探討[J].經(jīng)濟(jì)科學(xué),2012,(3):94-108.
[10]Anselin, L. Spatial econometrics: methods and models[M].Dordrecht: Kluwer Academic,1990.
[11]季民河,武占云,姜磊. 空間面板數(shù)據(jù)模型設(shè)定問(wèn)題探討[J].統(tǒng)計(jì)與信息論壇,2011,(6):3-9.
[12]SalaiMartin, X.Doppelhofer, G.Miller,R.I. Determinants of longterm growth: a bayesian averaging of classical estimates(BACE) approach[J].American Economic Review,2004,94(4):813-835.
[13]張望. 政府公共服務(wù)、產(chǎn)業(yè)集聚與經(jīng)濟(jì)增長(zhǎng)[J].山西財(cái)經(jīng)大學(xué)學(xué)報(bào),2012,(4):39-45.
(責(zé)任編輯:寧曉青)
A Study on Relationship Between Industrial Agglomeration and Provincial Economic Growth Based on Spatial Panel Econometrics
XIE Xiongjun,HE Hongqu
. (Business School, Central South University, Changsha, Hunan 410083,China).
Abstract:This paper used panel data of each provincial industrial agglomeration and economic development from 1985 to 2011 to do spatial econometric analysis. Studies include global spatial autocorrelation analysis, local spatial autocorrelation analysis, spatial panel lag model, spatial panel error model analysis respectively. Empirical results show that: China's industrial agglomeration and regional economic growth have obvious regional differences. There are significant spatial autocorrelation in both industrial agglomeration and regional economic growth. Industrial agglomeration has a positive spatial autocorrelation with regional economic growth, namely industrial agglomeration play a role in promoting regional economic growth.
Key words:Industrial Agglomeration; Provincial Economy; Spatial Econometrics; Spatial Panel Data; Moran's I
[6]錢曉燁,遲巍,黎波. 人力資本對(duì)我國(guó)區(qū)域創(chuàng)新及經(jīng)濟(jì)增長(zhǎng)的影響基于空間計(jì)量的實(shí)證探討[J].數(shù)量經(jīng)濟(jì)技術(shù)經(jīng)濟(jì)探討,2012,(4):107-121.
[7]Anselin, L., Florax, R. andRey, S. Advanced in spatial econometrics:methodology,tools and applications[M].Berlin:Springer Verlag,2004.
[8]陳建軍,胡晨光.產(chǎn)業(yè)集聚的集聚效應(yīng)以長(zhǎng)江三角洲次區(qū)域?yàn)槔睦碚摵蛯?shí)證探討[J].管理世界,2008,(6):68-83.
[9]王家庭,賈晨蕊. 我國(guó)城市化與區(qū)域經(jīng)濟(jì)增長(zhǎng)差異的空間計(jì)量探討[J].經(jīng)濟(jì)科學(xué),2012,(3):94-108.
[10]Anselin, L. Spatial econometrics: methods and models[M].Dordrecht: Kluwer Academic,1990.
[11]季民河,武占云,姜磊. 空間面板數(shù)據(jù)模型設(shè)定問(wèn)題探討[J].統(tǒng)計(jì)與信息論壇,2011,(6):3-9.
[12]SalaiMartin, X.Doppelhofer, G.Miller,R.I. Determinants of longterm growth: a bayesian averaging of classical estimates(BACE) approach[J].American Economic Review,2004,94(4):813-835.
[13]張望. 政府公共服務(wù)、產(chǎn)業(yè)集聚與經(jīng)濟(jì)增長(zhǎng)[J].山西財(cái)經(jīng)大學(xué)學(xué)報(bào),2012,(4):39-45.
(責(zé)任編輯:寧曉青)
A Study on Relationship Between Industrial Agglomeration and Provincial Economic Growth Based on Spatial Panel Econometrics
XIE Xiongjun,HE Hongqu
. (Business School, Central South University, Changsha, Hunan 410083,China).
Abstract:This paper used panel data of each provincial industrial agglomeration and economic development from 1985 to 2011 to do spatial econometric analysis. Studies include global spatial autocorrelation analysis, local spatial autocorrelation analysis, spatial panel lag model, spatial panel error model analysis respectively. Empirical results show that: China's industrial agglomeration and regional economic growth have obvious regional differences. There are significant spatial autocorrelation in both industrial agglomeration and regional economic growth. Industrial agglomeration has a positive spatial autocorrelation with regional economic growth, namely industrial agglomeration play a role in promoting regional economic growth.
Key words:Industrial Agglomeration; Provincial Economy; Spatial Econometrics; Spatial Panel Data; Moran's I
[6]錢曉燁,遲巍,黎波. 人力資本對(duì)我國(guó)區(qū)域創(chuàng)新及經(jīng)濟(jì)增長(zhǎng)的影響基于空間計(jì)量的實(shí)證探討[J].數(shù)量經(jīng)濟(jì)技術(shù)經(jīng)濟(jì)探討,2012,(4):107-121.
[7]Anselin, L., Florax, R. andRey, S. Advanced in spatial econometrics:methodology,tools and applications[M].Berlin:Springer Verlag,2004.
[8]陳建軍,胡晨光.產(chǎn)業(yè)集聚的集聚效應(yīng)以長(zhǎng)江三角洲次區(qū)域?yàn)槔睦碚摵蛯?shí)證探討[J].管理世界,2008,(6):68-83.
[9]王家庭,賈晨蕊. 我國(guó)城市化與區(qū)域經(jīng)濟(jì)增長(zhǎng)差異的空間計(jì)量探討[J].經(jīng)濟(jì)科學(xué),2012,(3):94-108.
[10]Anselin, L. Spatial econometrics: methods and models[M].Dordrecht: Kluwer Academic,1990.
[11]季民河,武占云,姜磊. 空間面板數(shù)據(jù)模型設(shè)定問(wèn)題探討[J].統(tǒng)計(jì)與信息論壇,2011,(6):3-9.
[12]SalaiMartin, X.Doppelhofer, G.Miller,R.I. Determinants of longterm growth: a bayesian averaging of classical estimates(BACE) approach[J].American Economic Review,2004,94(4):813-835.
[13]張望. 政府公共服務(wù)、產(chǎn)業(yè)集聚與經(jīng)濟(jì)增長(zhǎng)[J].山西財(cái)經(jīng)大學(xué)學(xué)報(bào),2012,(4):39-45.
(責(zé)任編輯:寧曉青)
A Study on Relationship Between Industrial Agglomeration and Provincial Economic Growth Based on Spatial Panel Econometrics
XIE Xiongjun,HE Hongqu
. (Business School, Central South University, Changsha, Hunan 410083,China).
Abstract:This paper used panel data of each provincial industrial agglomeration and economic development from 1985 to 2011 to do spatial econometric analysis. Studies include global spatial autocorrelation analysis, local spatial autocorrelation analysis, spatial panel lag model, spatial panel error model analysis respectively. Empirical results show that: China's industrial agglomeration and regional economic growth have obvious regional differences. There are significant spatial autocorrelation in both industrial agglomeration and regional economic growth. Industrial agglomeration has a positive spatial autocorrelation with regional economic growth, namely industrial agglomeration play a role in promoting regional economic growth.
Key words:Industrial Agglomeration; Provincial Economy; Spatial Econometrics; Spatial Panel Data; Moran's I
財(cái)經(jīng)理論與實(shí)踐2014年2期