Growth in a Cross-section of Cities: Location, Increasing Returns or Random Growth?

Abstract

This article analyses empirically the main existing theories on income and population city growth: increasing returns to scale, locational fundamentals and random growth. To do this we consider a large database of urban, climatological and macroeconomic data from 1,173 US cities observed in 1990 and 2000. The econometric model is robust to the presence of spatial effects. Our analysis shows the existence of increasing returns and two distinct equilibria in per-capita income and population growth. We also find important differences in the structure of productive activity, unemployment rates and geographical location between cities in low-income and high-income regimes.

Résumé

Cet article analyse de façon empirique les principales théories actuelles sur l'expansion des revenus et de la population dans les villes: augmentation des revenus à l'échelle, principes fondamentaux de la situation, et expansion aléatoire. Pour ceci, nous examinons une importante base de données urbaines, climatologiques et macroéconomiques relevées dans 1 173 villes des États-Unis de 1990 jusqu'à l'an 2000. Le modèle économétrique est solide en présence d'effets spatiaux. Notre analyse indique la présence de revenus en augmentation, et de deux équilibres distincts entre les revenus par personne et l'expansion de la population. Nous relevons également la présence d'importantes différences au niveau de la structure de l'activité de production, des taux de chômage, et de la situation géographique entre des villes à faibles revenus et à revenu élevé.

Resumen

Este artículo analiza de manera empírica las principales teorías existentes sobre el aumento de la renta y la población en las ciudades: rendimientos crecientes a escala, fundamentos de ubicación y crecimiento aleatorio. Para hacer esto consideramos una gran base de datos urbanos, climatológicos y macroeconómicos de 1.173 ciudades estadounidenses observadas en 1990 y 2000. El modelo econométrico es sólido ante la presencia de efectos espaciales. Nuestro análisis demuestra la existencia de rendimientos crecientes y dos equilibrios distintos en los ingresos per cápita y el incremento de la población. También encontramos diferencias importantes en la estructura de la actividad productiva, las tasas de desempleo y la ubicación geográfica entre las ciudades de regímenes de renta baja y alta.

作者: 摘要

本文实证分析了关于收入和城市人口增长的现有主要理论 : 按照规模、地点、基本面和随机增长几种要素，探讨收益增长状况。为此，我们根据1990 至 2000年获自 1,173 个美国城市的一个记载了城市状况、气候和宏观经济数据的大型数据库进行了分析。明确了计量经济模型与空间效应之间存在的十分缜密的关系。我们的分析结果表明，收益是递增性上升的，而且人均收入与人口增长之间有两种截然不同的平衡。我们还发现，低收入和高收入城市之间在生产活动的结构、失业率和地理位置方面存在着重要的差异。

Keywords:

• Threshold tests
• locational fundamentals
• multiple equilibria
• random growth

JEL Classification:

• C12
• C13
• C33
• O1
• R0
• R11

We had benefited from the helpful comments from David Cuberes, Fernando Sanz-Gracia, Elisabet Viladecans-Marsal and seminar participants at the Institut d'Economia de Barcelona (IEB). Earlier versions of this paper were presented at the 15th International Conference on Macroeconomic Analysis and International Finance (Crete, 2011), at the 65th European Meeting of the Econometric Society (Oslo, 2011), at the 58th Annual North American Meetings of the Regional Science Association International (Miami, 2011), at the XXXVI Symposium of Economic Analysis (Málaga, 2011), and at the Annual Conference of the International Association for Applied Econometrics (London, 2014), with all the comments made by participants being highly appreciated. Finally, suggestions and observations received from two anonymous referees and the editor (Prof. Bernard Fingleton) have also significantly improved the version originally submitted. All remaining errors are ours.

We had benefited from the helpful comments from David Cuberes, Fernando Sanz-Gracia, Elisabet Viladecans-Marsal and seminar participants at the Institut d'Economia de Barcelona (IEB). Earlier versions of this paper were presented at the 15th International Conference on Macroeconomic Analysis and International Finance (Crete, 2011), at the 65th European Meeting of the Econometric Society (Oslo, 2011), at the 58th Annual North American Meetings of the Regional Science Association International (Miami, 2011), at the XXXVI Symposium of Economic Analysis (Málaga, 2011), and at the Annual Conference of the International Association for Applied Econometrics (London, 2014), with all the comments made by participants being highly appreciated. Finally, suggestions and observations received from two anonymous referees and the editor (Prof. Bernard Fingleton) have also significantly improved the version originally submitted. All remaining errors are ours.

Notes

1. We use simpler notation, but this equation is equivalent to Eq. (2.7) and Eq. (1’) in Glaeser et al. (1995) and Glaeser (2000), respectively.

2. We acknowledge one anonymous referee for suggesting this point.

3. As a robustness check, we explore the suitability of alternative threshold models putting the emphasis on nonlinearities on the locational fundamentals rather than in the endogenous income and population variables, see Section 5.4.

4. There are important differences between Beaumont et al. (2003)'s data and our sample. Their sample size is 9 times smaller than ours and they consider regions from different countries, while our sample includes cities from only one country.

5. This is a macroeconomic approach to increasing returns. However, some of our exogenous variables, i.e. human capital variables, are considered in the literature to be a source of agglomeration economics from a microeconomic perspective (see Duranton & Puga, 2004). This micro-treatment of the model is beyond the scope of this paper.

6. There are 141 cities in our sample below the 25.000 inhabitants in 1990. As the sample is defined according to the largest cities in the latest period, it might imply a slight bias because these are the ‘winning’ cities, namely, those that have presented the highest growth rates (Black & Henderson, 2003). Nevertheless, as the period considered is only one decade, there are almost no ‘losing cities’ excluded that could bias our results. If we consider all the incorporated places with 25,000 or more inhabitants in 1990 according to the US Census Bureau only 10 out of these 1,077 cities fall below the 25,000 inhabitants in 2000.

7. In fact, information on most of the variables used in this study is only available at the place or metropolitan area level.

8. We also introduced state-level dummies into some of the preliminary estimations, but they were not significant.

9. These results are available from the authors upon request.

Additional information

Funding

This work was supported by the Spanish Ministerio de Economía y Competitividad [project numbers ECO2011-22650, ECO2013-45969-P and ECO2013-41310-R], the DGA (ADETRE research group) and FEDER.