Agglomeration and Congestion in the Economics of Ideas and Technological Change

American Journal of Economics and Sociology, The,  Jan, 2001  by Norman Sedgley,  Bruce Elmslie

• E-mail
• Print
• Link

ABSTRACT. Urban economists have long recognized that space is economically important. Evidence of the importance of urban agglomeration and the offsetting effects of congestion are provided in a number of studies of productivity and wages. Little attention has been paid to this evidence in the economic growth literature. The new growth research focuses on technological change. We extend the production function for new ideas common to this research in a way that allows for congestion and agglomeration in innovation and test the hypothesis that these forces are important in explaining innovation. Strong evidence is found that agglomeration and congestion are important in explaining the vast differences in per capita patent rates across US states. This suggests an important new agenda in linking studies of urban economics with the rapidly advancing field of endogenous growth.

I

Introduction

THE PURPOSE OF THIS PAPER is to extend the idea of agglomeration and congestion effects from regional and urban economics by drawing a link between this literature and the new models on the economics of ideas common to the new growth research. Empirical evidence from urban economics, regional economics, and economic geography suggests that agglomeration and congestion effects have an important impact on productivity differences across economies. While most of these studies measure the scale of production by the absolute size of the economy (the size of the population being most common) as many growth studies do, Ciccone and Hall (1996) are an exception. They show that population density across US States is important for explaining differences in productivity levels across states. Their analysis, however, suggests that agglomeration effects are much stronger than congestion effects. While their analysis is entirely static, it does suggest an important extension of the dynamic endogenous growth models.

Within the new endogenous growth theory there is a focus on the economics of ideas. This approach has proven useful to many wishing to understand the economics of innovation and technological change. These models, in their simplest form, suggest a scale effect: ceteris paribus a large economy innovates and grows at a faster rate than smaller economy. This unique implication is found to be at odds with the evidence (Helliwell and Chung, 1992; Jones, 1995a).

The absence of the scale effects suggested in these models concerns growth theorists. These scale effects tend to arise due to relaxed resource constraints and, therefore, a lower cost of innovation in a larger economy. An absence of these scale effects does not rule out the possibility of agglomeration and congestion having an important role to play in the new growth literature. The lack of evidence of scale effects, however, has caused the new growth literature to move in a direction that removes all notions of scale from the models. We believe new growth theory is overlooking some important and interesting evidence concerning the role of agglomeration and congestion in the production of new ideas and the dynamics of economic growth.

Our empirical analysis of agglomeration and congestion differs from past investigations. We use a direct measure of innovative outcomes as our dependant variable. Typically, researchers look for agglomeration and congestion in productivity measures. While these forces are well documented, little attention has been paid specifically to evidence of agglomeration and congestion in the production of new ideas. The production function for new ideas is at the heart of endogenous growth in much of the new growth research. The link between measures of innovation and productivity growth is well documented. Studies by authors such as Abramovitz (1986), Amable (1993), Fagerburg (1988, 1994), Verspagen (1991), Sedgley (1998), and Sedgley and Elmslie (2000) demonstrate that differences in innovation as measured by patenting statistics are important in explaining differences in productivity growth.

We give the density of economic activity an explicit role in a model where growth is driven by the production of new ideas and test the implications using state level data available from the US Federal Government. Our formulation allows us to gain some insight into how important agglomeration economies and congestion effects are for understanding the economics of ideas and innovation.

Section II reviews the related literature. Section III develops a theoretical reformulation of the production function for new ideas that allows for congestion and agglomeration in a straightforward way. Section IV describes the data collected for the research and empirical estimates are outlined in Section V. Section VI draws some final conclusions and suggestions for further research.

II

Related Literature

PACK (1994) CONVINCINGLY argues that there are very few true tests of endogenous growth theory. One of the salient implications of endogenous growth models is the prediction of scale effects, large economies innovate faster than small economies, ceteris pari bus. Some empirical studies look for scale effects, typically adding the size of the population to the list of explanatory variables within the neoclassical framework (See for example Helliwell and Chung, 1992). The evidence does not strongly support the existence of scale effects.