Small and medium-sized cities in the Age of the Urban Triumph Evert Meijers Delft University of Technology, Netherlands
The Age of the Urban Triumph The bigger, the better Agglomeration benefits are key to economic growth Explains rapid pace of urbanisation and rise of megacities
Agglomeration advantages Quantity and quality of urban functions (specialisation) Infrastructure Labour market matching More economic diversity more resilient Innovation, knowledge exchange City as production & consumption environment Elevator Greater variety in lifestyles / stadsluft macht frei Energy efficiency Household-optimisation Sharing, matching, learning are the mechanisms (Duranton & Puga, 2004)
The bigger, the better: labour productivity +5.8% (Melo et al., 2009)
Agglomeration disadvantages Exposure to environmental pollution / noise / a more stressful environment Crime Social deprivation Congestion Price/quality ratio houses High office rents and land prices Less social cohesion
Tide is turning
Tide is turning
Problem! Many European cities are small or medium-sized OECD: Lack of agglomeration benefits is biggest threat to competitiveness of smalland medium-sized cities
But, contradictory urban dynamics Urbanisation rate stable Large cities do not grow faster Large cities do not contribute increasingly more to economy Second-tier cities outperform capital-city regions Dynamics cannot be explained by conventional agglomeration theory!
A novel explanation: borrowing size & city network externalities A different interpretation of the geographical dimension of agglomeration economies
Urban network externalities
Urban network externalities Urban network externalities are defined as external economies from which firms and households can benefit by being located in agglomerations that are well embedded in networks that connect with other agglomerations (Capello, 2000; Burger & Meijers, 2016). Networks substitute for proximity (Johansson and Quigley, 2004). McCann and Acs (2011) argue that nowadays national and international urban connectivity is more important for urban performance than urban size. Camagni et al (2016) show that small cities manage to profit relatively more from a developed network of cooperative relationships with other cities than do large cities. Combination matters: local buzz & global pipelines (Bathelt et al., 2004)
metropolitan functions as proxy for agglomeration benefits Metropolitan functions Business related functions Financial services (banks) Advanced Producer Services Headquarters top-500 multinationals (revenues, employees) Conference/exhibition facilities Science related functions top-500 universities (Shanghai-index) international research co-operation Sports functions Seats in sports stadiums Important sports events (WC/EC) (summer) Olympic games (after 1948) Cultural functions Cultural venues (theatres, opera houses, galeries, museums) Cultural events (concerts, film festivals, art exhibitions) Index Metropolitan Functions Aggregate of subindices
Networks substitute for size A good network embeddedness in (inter)national networks (flows of people + information) significantly contributes to the presence of metropolitan functions in cities Still, size is on average 2.5x more important than network embeddedness But for particular metropolitan functions network embeddedness is more important than size
Relative importance size and networks for individual metropolitan functions Local Size Regional Networks (Inter)national Networks International Organisations + 0 ++ Headquarters + 0 ++ Banks + 0 ++ Advanced Producer + - ++ Services Fairs + 0 ++ Universities ++ 0 0 Conferences + - ++ Patents + - ++ Cultural Venues ++ - + Cultural Events ++ 0 0 Sport Venues ++ 0 0 Sport Events ++ - 0
But Importance of size, regional network embeddedness and international network embeddedness differs for different sizes of cities
Figure 1: Fitted equation and 95% confidence interval for metropolitan function index and local population size (average marginal effect) An increase in city size has a minor effect on the presence of metropolitan functions in large cities (>200k), but a large effect for smaller cities.
Figure 3: Fitted equation and 95% confidence interval for metropolitan function index and (inter)national network embeddedness by local population size (average marginal effect) An increase in (inter)national network embeddedness is positive for all cities, but relatively more positive for larger cities
Figure 2: Fitted equation and 95% confidence interval for metropolitan function index and regional network embeddedness by local population size (average marginal effect) An increase of regional network embeddedness is positive for larger cities (>400k), not necessarily for smaller cities (agglomeration shadows?)
Regional network embeddedness: a closer look where relevant?
Polycentric Urban Regions in Europe
Urban network externalities through integration in polycentric metropolitan regions
Indicators functional integration
Indicators Institutional integration
Indicators cultural integration
Finding: importance of integration The stronger the cities in PURs are functionally integrated, the better their performance (more agglomeration economies) Institutional integration, or metropolitan governance, has a positive effect on the performance of PURs Effect is smaller than for functional integration Most important is that there is some form of metropolitan cooperation, but its exact shape and power/autonomy seem of secondary importance. There is some indication that the duration of cooperation plays a role, with longer lasting networks somewhat associated with better performance; Mixed evidence for cultural integration The main challenge is to move from fragmentation to integration, which demands a larger process of regionbuilding referred to as metropolisation, in which the economic, functional, administrative and socio-spatial qualities and features once attributed to the 'city' are reconstructed by citizens, firms and institutions at the scale of the PUR.
Ongoing work: development of a new city network indicator Based on co-occurrence of toponyms in text corpora, in this case the Web Archive CommonCrawl (all 25 million.nl websites) Scientometrics: The use of co-occurrence data in scientometrics assumes that the greater the probability of two elements co-occuring in the same article, the more strongly they are related. (Chavalarias and Cointet, 2013)
The co-occurrence method: existing application in spatial sciences Tobler and Wineburg, 1971 Devriendt et al. 2008
The Dutch urban system according to the Web absolute value
The Dutch urban system according to the Web relative value
Places with a strong network position (left) and those with a weak network position (right) in the Netherlands. Plaatsen met een zwakke en sterke positie in (nationale) netwerken Sterk verbonden plaatsen % Minder geïntegreerde plaatsen % 1. Roermond 20,11 91. Capelle aan den IJssel -15,62 2. Middelburg 16,71 92. Spijkenisse -14,10 3. Zutphen 13,32 93. IJsselstein -10,74 4. Maastricht 12,01 94. Landgraaf -10,72 5. Zwolle 10,40 95. Hellevoetsluis -9,70 6. Hoogeveen 10,00 96. Vlaardingen -9,65 7. Gorinchem 9,82 97. Zwijndrecht -9,11 8. Wageningen 9,75 98. Almere -8,92 9. Vlissingen 9,34 99. Etten-Leur -8,80 10. Alkmaar 8,84 100. Kerkrade -8,40 17. Delft 7,76
Borrowed size
The Concept of Borrowed Size Alonso, 1973: Borrowed size describes and explains the situation that especially smaller cities that are located in a larger megapolitan complex do perform better because they have access to ( borrow ) the agglomeration benefits of larger neighbouring cities. Meijers & Burger (2016): not necessarily a onedirectional process between a smaller and a larger city, but between cities that are close and within each other s sphere of influence Borrowed size refers to: a.) functions b.) performance
The other side of the medal: agglomeration shadows Cities cannot just borrow size from other cities in the urban system, they also face competition effects agglomeration shadows Growth near (higher tier) agglomerations will be limited due to competition effects (Dobkins & Ioannides, 2001). Agglomeration shadows play a negative role on growth of close-by small U.S. metropolitan areas (Partridge et al., 2009).
Randstad
Ongoing work: understanding why some cities borrow size and others face an agglomeration shadow Dimensions of borrowed size Connection size function Less functions As many More functions than expected functions as than expected given size expected given size Connection size performance Performs less than expected given size Performs as expected Performs better than expected given size 8 7,5% 17 16% 8 7,5% 14 13,2% 18 17% 11 10,4% 16 15,1% 11 10,4% 3 2,8%
Cities in the agglomeration shadow Gouda Veenendaal Soest Vlaardingen Krimpen aan den IJssel Velsen Schiedam Maassluis
Social Status
Historical importance
Household Income levels
Conclusion Dominance of urban triumph narrative focusing on agglomeration benefits as key driver of growth, prosperity and even sustainability has directed attention away from small and medium-sized cities In many parts of Europe, the actual urban dynamics suggest that also small- and medium-sized cities can do well My explanation is that agglomeration benefits are still important, but not geographically confined to large metropolitan areas: cities can borrow size in networks on a variety of spatial scales. This opens up a new range of opportunities for small and medium-sized cities to organise more agglomeration benefits for their firms and citizens.
Köszönöm! Thank you very much for your invitation and attention! Evert Meijers e.j.meijers@tudelft.nl