論文：空間組織與經(jīng)濟(jì)增長(zhǎng)

論文：空間組織與經(jīng)濟(jì)增長(zhǎng)

摘要：
生產(chǎn)要素和技術(shù)在地理空間上的分布與組織對(duì)生產(chǎn)率增長(zhǎng)是十分重要的。本文試圖建立一個(gè)經(jīng)濟(jì)增長(zhǎng)的空間模型：在連續(xù)空間上知識(shí)溢流與生產(chǎn)空間結(jié)構(gòu)通過創(chuàng)業(yè)組織相互促進(jìn)。在這個(gè)模型中，創(chuàng)業(yè)者促進(jìn)經(jīng)濟(jì)增長(zhǎng)主要通過兩個(gè)途徑：一方面它選擇具有優(yōu)勢(shì)的區(qū)位并開發(fā)那里的聚集效益（包括知識(shí)溢流），另一方面，它采用有效的內(nèi)部組織結(jié)構(gòu)將主意轉(zhuǎn)變成產(chǎn)品。為了使地理結(jié)構(gòu)成為生產(chǎn)的一部分內(nèi)容，本文將生產(chǎn)要素和技術(shù)的空間范圍引入生產(chǎn)函數(shù)，假定與生產(chǎn)有關(guān)的技術(shù)的數(shù)量和區(qū)位是已知的。本文提出的空間增長(zhǎng)模型顯明，一個(gè)經(jīng)濟(jì)體的長(zhǎng)期增長(zhǎng)不僅取決于生產(chǎn)要素的增長(zhǎng)和技術(shù)變化，還取決于它們?cè)诳臻g上的結(jié)構(gòu)擴(kuò)張。經(jīng)濟(jì)增長(zhǎng)是空間積累與組織的過程，伴隨著生產(chǎn)的地理擴(kuò)散與密集。

關(guān)鍵詞：空間分布；經(jīng)濟(jì)增長(zhǎng)；產(chǎn)業(yè)組織；模型

研究領(lǐng)域：區(qū)域經(jīng)濟(jì)學(xué)，產(chǎn)業(yè)組織理論

Spatial Organization and Economic Growth[ This paper was presented at the 54th Annual North American Meeting, Regional Science Association International, November 16-18, 2006, Toronto, Canada. One earlier version of this paper was circulated as “A Spatial Model of Growth: Taking Technology Seriously,” Ma* Planck Institute of Economics Discuss Paper #2006-12. The author is grateful to Giuseppe Arbia and Jagannadha P. Tamvada for their helpful comments on earlier drafts of the paper.]
Zuoquan Zhao
Ma* Planck Institute of Economics, Jena, Germany
(July 2007)

Abstract
The way factor inputs and technology are distributed and organized over geographical space accounts for productivity change. This paper attempts to develop a simple model of economic growth in continuous space in which knowledge spillovers and geographical organization of production reinforce each other through entrepreneurial organization. In our model, growth is enhanced by profit-seeking entrepreneurs who choose an advantageous location to e*ploit local agglomeration economies (including knowledge spillovers), and adopt an efficient internal organization to transform ideas into products. To internalize geographical structure, including spatial knowledge spillovers, we introduce the spatial ranges of factor and knowledge inputs into the production function under the assumption that the number and location of technologies that are embodied into products are observable or countable. Our model shows that the long run growth of an economy is determined not just by increase of factor inputs and technology but by their e*pansion in spatial structure. Growth is a process of spatial accumulation and organization: characteri
……（新文秘網(wǎng)http://m.120pk.cn省略3303字，正式會(huì)員可完整閱讀）……　
ge spillovers and spatial organization combined present an integrated e*planation of growth.
This research is motivated by the relationship between knowledge spillovers and spatial structure and by the role of entrepreneurial organization in production in geographical space. Empiric evidence shows that knowledge spillovers and spatial structure reinforce one another in geographical space (Audretsch and Feldman 1996; Caniels and Romijn 2005); knowledge spillovers are geographically bounded and limited by the spatial structure of economic activity (Jaffe et al. 1993; Feldman 1999). On the one hand, technology diffuses across locations or cities with factor mobility and trade over space. On the other hand, agglomeration of available technological ideas leads to faster generation of new ideas because many new ideas originate from a combination of old ones (Weitzman 1998). This suggests knowledge spillovers do not take a random walk in space. In fact, it is the profit-seeking entrepreneur who takes advantage of local knowledge spillovers (Lucas 1988; Schmitt 1989; Segerstrom 1991; Audretsch and Lehmann 2005) and other location-specific agglomeration benefits, e.g. easy access to market and supply (Krugman 1991a) and dynamic efficiency (Porter 1994; Ketelhöhn 2006). The entrepreneur begins and runs business not only with the choice of location (and relocation) but of organization, a factor long speculated as a fourth factor of production (Marshall 1890; Harbison 1956). Firms grow with organization capital, the accumulation of firm-specific knowledge (Prescott and Wisscher 1980). For firms, their internal organization accounts for economic performance and efficiency (Chandler 1962, 1977; Marris and Mueller 1980; Williamson 1981). Atkeson and Kehoe 2005); transaction costs would be saved if firms limit the spatial distribution or reduce the geographical range of production (Coase 1937). The role of organizational innovations in economic development is well addressed by Schumpeter (1934), while the ideas about how factor and knowledge inputs are organized in production receive little attention in endogenous growth theory.
To e*amine the role of both knowledge spillovers and spatial structure, we propose a spatial production function by making several modifications to the production function. First, we take into account the number of technologies that are embodied into products by firms regardless of their origins or sources (innovation, imitation, or learning). We follow the assumption by Romer (1986) that the new knowledge created by one firm “cannot be perfectly patented or kept secret,” and thus can be used simultaneously by many other firms. Thus, one individual technology is counted as many times as the number of the firms which employ this technology to make products at the same time. E*amples of this kind include Johnston (1966) and Gort and Konakayama (1982).
Second, we measure spatial structure and its intensity by taking into account its 2-D range and benchmarking against spatial randomness, a distribution in which there is no e*ternality or interaction among individuals of agents (Papageorgiou and Smith 1983). We measure spatial range in terms of the (weighted) mean distance that indicates the degree of dispersion of economic activity from its center of gravity, a spatial reference point first used by Walker (1874) who identifies the center of spatial distribution of U.S. population. Then we measure the e*tent of spatial intensity by using the ratio of the amount to the geographical range of economic activity[ Martin and Sunley (2003) suggest that geographical range should be considered in the measurement of agglomeration of economic activity.]. This inde* can identify the scope of agglomeration of an economy (local or territory-wide) in comparison with the spatial range of the territory underlying the economy. Unlike the widely used Gini coefficient and the Ellison-Glaeser inde* (Krugman 1991b; Audretsch and Feldman 1996; Sweeney and Feser 2004), our inde* can accurately calibrate the e*tent of spatial compactness. It allows for spatial heterogeneity both within and among firms regardless of the scale of geographical space[ The literature pays more attention to the spatial heterogeneity of agents and less to the spatially heterogeneous structure of firms (Maskell 2001; Plummer and Sheppard 2006).]. It facilitates the aggregation of economic activity of varying size across locations without loss of information. In particular, our intensity inde* is consistent with the principle of spatial interaction that, other things equal, interaction among individuals decreases with distance (Papageorgiou and Smith 1983).
Third, we introduce the geographical range of factor inputs and technology to the production function, relating production agglomeration to growth. The spatial production function illustrates that output is a function of factor and technology inputs and of their spatial structure. It shows that the realization and nature of scale economies depends upon how factor and technology inputs are organized. Thus, the content of scale economies and replication are not well applied to spatial production analysis due to the presence of technology and spatial structure in the production function.
Therefore, fourth, we provide two individual-based mechanisms of factor agglomeration and matching to address the role of entrepreneurial organization in production over space. We e*tend the scope of agglomeration economies to a spatial continuum with one e*treme the firm and the other e*treme the economy in a bounded space. We assume that e*ternalities not only e*ist among individual agents over space as showed in Papageorgiou and Smith (1983) and Papageorgiou (1978) but also appear among individual components of physical capital and technology. This e*tension is based on the observation that many new ideas originate from integrating old ideas and these new ideas that embody into products in a firm can be imitated by other firms. It suggests that a key role an entrepreneur plays in production is to internalize a variety of individual-level e*ternalities; and the efficiency of production in a firm depends upon the tradeoff between agglomeration economies against agglomeration diseconomies over geographical space. Thus we use agglomeration economies and factor matching to e*plain growth, complementing the notions of scale economies and factor substitution. Agglomeration economies, for e*ample, apply to an entire economy in space when spatial structure e*pands with knowledge spillovers, factor mobility and trade across all cities in the economy. In the literature, agglomeration-related economies or e*ternalities are divided into many categories (Wood and Parr 2005; Papageoigiou and Pines 2000; Glaeser et al. 1992) and their scope is mostly limited to a single location like cities or to smaller areas like states and regions in a country (Ciccone and Hall 1996; Fujita and Thisse 2000; Duranton and Puga 2003; Rosenthal and Strange 2004).
Our model is consistent in spirit with two lines of studies. One links spatial agglomeration with growth (Martin a ……（未完，全文共51582字，當(dāng)前僅顯示9278字，請(qǐng)閱讀下面提示信息。收藏《論文：空間組織與經(jīng)濟(jì)增長(zhǎng)》）