Trade, innovation and agglomeration. A case study for Colombia/Comercio, innovacion y aglomeracion. Un caso de estudio para Colombia/Comercio, inovacao e aglomeracao: um estudo de caso para a Colombia.
| Autor | Guevara-Rosero, Grace Carolina |
| Cargo | Research article texto en ingles |
-
Introduction
Spatial disparities arise not only between countries but regionally within them. This process of polarization merits special attention. Among the determinants, technological progress has been identified since Lucas' 1988 study as an engine of economic growth. A region that pursues investment in R&D can benefit from productivity gains, attracting more economic activity and leading to a process of industrial agglomeration. Such a process goes hand in hand with the increased population density, resulting in higher demand (Krugman, 1991). Regions with larger populations are attractive due to their large local markets and a wide variety of products, which in turn attracts more people at the expense of smaller regions. This picture might be shaped by the influence of trade globalization. With trade, regions can have the same variety of products and even services, making them more homogeneous. In this sense, trade might reduce inter-regional disparities. However, products still enter countries through specific ports of entry, generally in large cities, reinforcing their prevalence in the national spectrum. The effect of transport costs has been extensively studied in the New Economic Geography (NEG) literature. This factor shapes the spatial distribution of economic activities because lower costs facilitate inter-regional trading. According to Krugman (1991), once transportation costs are low enough, population concentrates in certain regions, increasing divergence between them. Firms determine production distribution factors depending on the scale of transport costs (Krugman, 1991; Krugman & Elizondo, 1996; Krugman & Venables, 1995; Venables, 2005; Monfort & Nicolini, 2000; Rauch, 1991; Alonso-Villar, 1999; Paluzie, 2001).
This paper focuses on the relationship between the level of population agglomeration within regions and innovation, and how imports affect this relationship. The main aim is to identify the source of trade that drives the agglomeration effect of innovation, by distinguishing the type of trade according to technological intensity.
This study attempts to bridge two factors that have been separately studied separately: trade on the one hand; science, technology and innovation activity (STI), on the other. The mechanism explaining this intuition is that foreign technological knowledge transmitted by imports increases domestic technology stock, in turn raising domestic productivity (Keller, 2004; Eaton & Kortum, 2006). Greater efficiency is reached in places where exporting firms operate with economies of scale (Esfahani, 1991; Balassa, 1978). In addition, trade is a channel of technology diffusion between developed and developing countries as goods that incorporate technology can be imitated or adapted in the receiving country (Coe & Helpman, 1995). Nevertheless, productivity gains are not uniformly diffused throughout the national territory. Particular regions with economic advantages may benefit the most. Thus, the presence of manufacturing and skilled labor becomes increasingly concentrated in core regions. This phenomenon has been verified in the case of China (Ge, 2006; Kanbur & Zhang, 2005). Likewise, Henderson and Kuncoro (1996) demonstrate that in Indonesia, manufacturing industries concentrate in large metropolitan areas as trade liberalization increases. In the case of Mexico, trade liberalization encouraged the establishment of manufacturing activities in Northern Mexican states near to the United States (Hanson, 1997; Madariaga et al., 2004; Chiquiar, 2005; Aroca et al., 2005; Jordaan & Rodriguez-Oreggia, 2012). This may be an indication that technology diffusion through trade is more effective at shorter distances.
To conduct this study, regional trade data from Colombia was used. This type of data allowed us to assess for the first time the influence of regional imports on the intraregional agglomeration-innovation relationship, taking into account the technological intensity of goods involved. The distinction between the inter-regional and intra-regional levels is not trivial. Imports affect regions differently; as tariffs vary depending on whether goods are derived from regional specialization (Head & Mayer, 2004). Interregional configurations are likely to be modified by imports. Further, intra-regional configurations are also affected as agglomeration forces operate at the local level (Head & Mayer, 2004). Cities with an already high concentration of the population can be enhanced by imports (Hansen, 1990). Therefore, the study of agglomeration within regions is theoretically and empirically well-grounded.
Panel data from 32 regions over the period 2000-2009 was used, with unobserved heterogeneity being controlled using the appropriate techniques. Unobserved heterogeneity is given by the fact that patterns of concentration depend on regional characteristics that are not always observable. The fixed effects model was then employed using a LSDV (Least Square Dummy Variable) estimator with robust standard errors allowing for intra-group correlation.
Concerning regional imports, the average annual growth rate was 4% in the period studied. This increasing level of imports is related to population agglomeration within Colombian regions with a significant correlation of 55%. More importantly, the composition of imports seems to be of great importance in such a relationship. Looking at the correlation coefficients, primary product imports and imports with low-technological intensity are highly correlated with increased population concentration (62% and 53%, respectively) while hi-tech imports are less correlated with the agglomeration pattern (46%). This means that when imports enter a region, local firms have opportunities to innovate using the technology embedded in those imported goods. Other firms are then attracted to that region and agglomeration continues.
This paper is organized as follows. Section 2 reviews the literature on trade, innovation and spatial concentration. Section 3 describes stylized facts regarding regional concentration, imports according to technological intensity and levels of science, technology and innovation in Colombia. Sections 4 and 5 present the data and methodology. Section 6 discusses the results, and section 7 then concludes.
-
Theoretical framework of trade, innovation and spatial concentration
As economies liberalize in terms of trade, they specialize under their comparative advantages. This Ricardian prediction results in a pattern of lagged and advanced countries since the former are commonly specialized in low-technology products whereas the latter specialize in high-technology products. Notwithstanding, trade openness has been seen as a facilitator of technology diffusion across space and time. Thus, countries lagging in terms of technology might benefit from better access to technical advances developed around the world (Mayer, 2000).
At first, it was established that technical change is an endogenously produced outcome (Grossman & Helpman, 1991; Romer, 1990; Aghion & Howitt, 1992), with scholars pointing out the role of innovation as an engine of growth. Other theorists went beyond explaining the mere existence of technology and focused on its diffusion. The pioneers Eaton and Kortum (1999) considered a model of several countries producing output while using combined inputs under constant returns to scale. For them, new technologies are the result of R&D investment, and innovation becomes productive only if it is diffused. The main finding was that technology spillovers increase research productivity in other places. New technology carried out in specific locations may benefit other locations, with trade playing a technological dissemination role (Grossman & Helpman, 1991). In Eaton and Kortums' 2002 model, trade allows access to a wide variety of foreign goods, and therefore access to foreign production technologies.
According to theoretical models, the interaction between domestic and foreign firms through international diffusion of technology increases domestic productivity. This is to say that foreign technological knowledge transmitted via imports increases domestic technology stock, and in turn raises domestic productivity (Keller, 2004; Eaton & Kortum, 2006). Santacreu (2015) explains the connections between trade and growth, showing that non-innovative countries benefit from technology spillovers through imports, leading to a reduction in the innovation gap between themselves and their trading partners. She found that 90% of growth in Asia is explained by imports from the USA and Japan. Other studies have focused on the technology-diffusing effect of FDI (as opposed to imports) on productivity. Robust estimations show positive technological spillovers from FDI (Xu, 2000; Keller & Yeaple, 2009).
The role of trade in relation to innovation and the productivity level of firms is emphasized. Nevertheless, the implications of that relationship on the distribution of economic activity is still in question. This can be analyzed by turning to New Economic Geography. The standard finding in this branch of economics is that trade integration increases regional concentration of the economic activity depending on their initial characteristics. The location with the largest market size, or population, will host the majority of local firms (Krugman, 1980; Martin & Ottaviano, 1999). Geographical characteristics also matter. Trade increases agglomeration in geographically-advantaged regions to the detriment of disadvantaged regions (Nishikimi, 2008; Alonso-Villar, 1999; Crozet & Koenig, 2004). In addition to inter-regional spatial configuration, trade also affects the intra-regional distribution of economic activity in cities. Henderson (1982) and Rauch (1991) examined the distribution of cities in an open economy through an analysis of urban systems. The hypothesis of Rauch's trade-urban model (1991) is that the cities...
Para continuar leyendo
Solicita tu pruebaCOPYRIGHT GALE, Cengage Learning. All rights reserved.
