Smartness and European urban performance: assessing the local impacts of smart urban attributes

Abstract

In this paper we adopt a comprehensive definition of Smart City and examine whether and how the main characteristics of urban smartness are growth-enhancing. With a sample of cities from the European Union we assess the city-specific impacts on urban performance of a complex “urban smartness” indicator by applying a Spatial Autoregressive Local Estimate to an urban production function. The results of this econometric analysis point in the direction of the importance of space-specific characteristics in shaping the economic effect of smart urban qualities, providing grounding to place-based public policies that account for local characteristics. We also identify different clusters with respect to the impacts of smartness on urban performance and wealth, highlighting the need for geographically differentiated policy actions.

Keywords:

• smartness
• urban performance
• cities

JEL classification codes:

• O18
• O43
• O44

Acknowledgement

Chiara Del Bo gratefully acknowledges the following research grant: Dote Ricercatori: FSE, Regione Lombardia.

Notes

1. UN (2007).

2. A complete list of the cities in our sample is available in Appendix A1.

3. See Glaeser (2008) for a recent overview with a perspective based on the concept of spatial equilibrium, and Camagni (1993) for a broader vision on the determinants of urban performance.

4. For a critical review of the concept of localized knowledge spillovers, see Breschi and Lissoni (2001) and Capello (2009).

5. An interesting applied perspective on the notion, and measurement, of smart cities has been provided by the privately funded research project “Smart cities – ranking of European medium-sized cities”. For more information on the project findings, see Giffinger et al. (2007).

6. The use of the number of visitors to a museum within the Smartness indicator can be related to the notion of usage intensity, in line with traditional infrastructure efficiency studies. This measure is thus different from the pure museum availability for the sampled cities, which can instead be viewed as an amenity measure (thus properly proxying one element in the benefit function of the analyzed cities).

7. See Organization for Economic Cooperation and Development (2005) for further details. Knowledge Intensive Services are henceforth abbreviated as KIS.

8. This is in particular true for the Interreg IVB North Sea Region Programme project “Smart Cities” (www.smartcities.info), which deals in particular with the assessment of the contribution of e-government to urban development. The theoretical paper to which we refer and some of the ideas developed in the present article have been developed within this project.

9. The number of cities included in this analysis is mainly motivated by the data constraint issue. We aimed to cover the widest possible set of cities with the Urban Audit data set while at the same time covering all indicators listed in Table 1.

10. Mathematical details are given in Appendix A2.

11. Pearson's correlation index equals 0.46, significant at all conventional levels, between the log of GDP and the smartness indicator.

12. The low levels of the local impact of smartness may be partially due to the use of the SALE model, which is based on a geographical notion of space, associated with the characteristics of our sample.

13. We thank an anonymous referee for suggesting us this logical implication of our applied analysis.