Presentation by Antonio Accetturo, Head of the Economic Research Unit at the Bank of Italy – Trento branch, at the second meeting of the Spatial productivity Lab of the OECD Trento Centre held on 4 December 2018. More info http://oe.cd/SPL

1. BANCA D’ITALIA
E U R O S I S T E M A
Antonio Accetturo
Spatial Productivity Lab Meeting
Trento, December 4th, 2018
Urban Development,
Agglomerations, and Aggregate
Growth in Italy

2. A Bank of Italy research project
• Coordinators: A. Accetturo, A. Lamorgese, S. Mocetti, P. Sestito
• From 2015 to 2018
• 20 papers
• Workshops:
• 2 intermediate (dec. 2015, dec. 2016)
• 1 final (june 2018)

3. Urban areas and growth
• Cities are the engines of growth in advanced economies (Glaeser,
2011). Urban areas are more productive (Melo et al., 2009), more
innovative (Duranton and Puga, 2001), and guarantee a more
efficient allocation of the workforce (De La Roca and Puga, 2017).
• The importance of urban areas has grown in the last 100 years
• Modern economies are inherently urban. There is a strong correlation
between growth and agglomeration (Castells-Quintana, 2017; Frick e
Rodriguez-Pose, 2018)
Country
Decadal growth
PeriodUrban pop. Total pop.
Italy 7.2% 4.9% 1911-2001
United States 17.9% 12.6% 1920-2010
Spain 18.1% 9.0% 1920-2010
France 7.7% 5.7% 1937-2007

4. Italy is one of the cradles of urban
civilization in Europe
In the late Roman Empire there were 2.5 cities or villages per 1000
sqkm, against 1.2 in France, 1.0 in Spain and 0.5 in Germany and UK

5. Italy today: a much more limited role for
large urban agglomerations
• In cities with at least 500k inhabitants there is 31% of Italian
population; 38% in Spain, 40% in Germany, 41% in France,
and 41% in UK
0%
10%
20%
30%
40%
50%
60%
Francia Germania Italia Regno Unito Spagna
Parigi
Monaco
Milano
Londra
Madrid
Marsiglia
Berlino
Roma
Manchester
Barcellona
Source: Eurostat; histograms indicate the share of value added produced in each (group of) urban areas; dark blue: first
urban area; blue: second urban area; light blue: al the cities between the 3rd and the 5th; white: all urban areas between
the 6th and the 10th.

6. The structure of the project
Urban System Implications for
growth
Historical origins
Agglomeration
benefits
Agglomeration
costs

7. The structure of the project
Urban System Implications for
growth
Historical origins
Agglomeration
benefits
Agglomeration
costs

8. The origins of an Urban System
Literature has focused on two main determinants:
• Geographical Characteristics
• Historical Shocks
• Geographical characteristics (orography, geology,
seismicity) influence on the ability of an area to sustain
large population
• However, la geography explains a small portion of the
heterogeneity in the spatial distribution of population
(Combes et al., 2010; Bleakley e Lin, 2015)
• Italy is no exception: geographical characteristics explain
30% of the city size distribution

9. The origins of an Urban System
Literature has focused on two main determinants:
• Geographical characteristics
• Historical shocks
• Dincecco and Onorato (2016): frequent conflicts in Northern
Italy (IX-XVIII cen.) generated urbanization (safe harbor
effect)
• Cervellati et al. (2018): positive correlation between the size
of a country and city growth (especially capitals) in the
period XI-XIX cen.
• Accetturo, Cascarano, de Blasio (2018a): the role of Pirates’
attacks (IX-XIX cen.) in Central and Southern Italy;
population concentrated in less accessible areas (more
rugged and distant to the sea), hampering urban
development

10. Long run effects
“Urban Systems are remakably stable” (Hohenberg, 2004):
• In the North: several medium-small cities (competition
and cannibalization)
• In the South: two large (parasitical? Bosker et al., 2008)
administrative centers and the rest of the population
dispersed in inaccessible areas
• Starting from the 1950s’: change in the distribution of
population (due to structural transformations:
industrialization and servitization)

11. Dynamics of population: 1951-2011

12. Why did larger urban areas grow less?
Starting from 1970s’, smaller relative grow rates for larger
urban areas are due to several reasons:
• Industrial districts and the “Third Italy”
• Increasing congestions costs in larger metropolitan areas
• Lamorgese and Petrella (2018) show larger commuting
areas (CAs) grew by incorporating neighboring smaller
CAs (from 1981 to 2011 the total number of CA has
decreased from 954 to 612).

13. The enlargement of urban CA

14. The structure of the project
Urban System Implications for
growth
Historical origins
Agglomeration
benefits
Agglomeration
costs

15. Congestion costs
• Traffic congestion in Italian cities is as large as in France and
larger than in Germany and Spain (but cities in Italy are smaller)
• Loschiavo (2018): living in a bigger city is correlated with a
smaller life satisfaction (due to commuting time)

16. Housing prices in the long run
• Strong increase in housing prices from the 1970s’
• Italy’s growth driven by construction costs and, especially,
land prices
0,0
0,5
1,0
1,5
2,0
2,5
3,0
1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010
prezzi delle case, 1970=1
Media di 14 paesi Italia
Source: Cannari, D’Alessio e Vecchi (2016), I prezzi delle abitazioni in Italia, 1927-2012, Banca d’Italia, Occasional
papers; Casolaro e Fabrizi (2018), House prices in local markets in Italy: dynamics, levels and the role of urban
agglomerations, Banca d’Italia, Occasional papers, Schularick e Steger (2018), No price like home: global house prices,
1870-2012, American Economic Review.

17. Source: Andrews, Caldera-Sánchez e Johansson (2011), Housing markets and structural policies in OECD Countries, OECD
working papers
• The elasticity is on average low but highly heterogeneous
• It may depend on both geographical and regulatory
reasons
0,0
0,5
1,0
1,5
2,0
2,5
CHE NLD AUT ITA BEL FRA ISR GBR DEU POL ESP NOR AUS IRL NZL FIN JPN CAN DNKSWEUSA
The elasticity of housing supply

18. CHENLDAUT ITA
BEL
FRA
GBRDEU POL
ESP NORAUS
IRL
NZL
FIN
JPN
CAN DNK
SWE
USA
ISR
0,0
0,5
1,0
1,5
2,0
2,5
0 50 100 150 200 250
Elasticitàdell'offertadicase
giorni per ottenere un permesso di costruire
coefficiente di correlazione: -0.56***
• The elasticity is larger in countries with less regulation;
the same is tru within Italy
• Geography matters but not much
Fonte: Andrews, Caldera-Sánchez e Johansson (2011), Housing markets and structural policies in OECD Countries, OECD
working papers.
Determinants of housing supply

19. Fonte: Manzoli e Mocetti (2016), I prezzi delle case e il gradiente centro-periferia: evidenze dalle
città italiane, Questioni di Economia e Finanza.
• Strong differences between North and South, urban and
non-urban, and, especially, centers and peripheries
• Localization (e.g. proximity to the center) is the most
important determinant for housing prices: differences
among centers are more marked that differences among
peripheries
Housing prices within Italy
prezzidellecase

20. The Center-Periphery gradient

21. Better transportation infrastructures are likely to expand the
commuting areas with small impacts on commuting times.
This reduces the centripetal pressures on city centers (see,
the impact of the new light railway in Florence, Budiakivska e
Casolaro, 2018)
Source: Manzoli and Mocetti (2016), I prezzi delle case e il gradiente centro-periferia: evidenze dalle città italiane, Occasional paper.
The impact of transportation
infrastructures

22. The structure of the project
Urban System Implications for
growth
Historical origins
Agglomeration
benefits
Agglomeration
costs

23. Huge literature on urban premia in terms of
productivity, wages, and innovation:
o Is Italy somewhat different?
o How large are these premia compared with
other countries?
o Why do we observe differences?
Agglomeration benefits

24. • VA per worker is larger in Italian urban areas with
respect to non-urban areas: ~30%, constant from 2005
to 2014
• Not a small number w.r.t. other countries
• Productivity premium reduces to 1/3 if we control for
sectoral composition and firm size
• Sorting and agglomeration explain the urban productivity
premium; the magnitude of sorting is larger than the one
for agglomeration (Lamorgese e Petrella, 2018)
Productivity

25. Cascarano, D’Addario, Lamorgese (2017) find that:
• Per capita patent applications are larger in urban areas:
– 1.3 active inventors per 1,000 inhabitants per year (0.7 in non-urban areas);
0.7% firms per year apply for a patent (0.3% in non-urban areas)
• Differently from other countries (e.g. Carlino for the US) no
wage premia for inventors; there is a wage premium for all
workers of innovative firms.
Innovation

26. • Lamorgese, Olivieri e Paccagnella (2018): estimation of
urban wage premia
– International consensus: 3-8% (2-5% controlling for
composition effects)
– Italy: 4,5% (0,5% controlling for composition effects)
• Possible explanations:
– Nation-wide wage setting
– Non-monetary amenities (but Lo Schiavo, 2018 says the
opposite)
– Low mobility of workers (also due to low elasticity of
housing supply)
Wages

27. The structure of the project
Urban System Implications for
growth
Historical origins
Agglomeration
benefits
Agglomeration
costs

28. • History and geography (and their interactions) shape
the Italian urban system
– Polycentric system and lower urban concentration
– Until WWII high dispersion of population (in the South in
inaccessible areas)
• Braked Urban Growth
– Large urban centers stopped growing in the 1970s’
– Growth by enlarging commuting areas (without major
infrastructural investments)
• High congestion and housing prices
– Steep center-periphery gradient;
– In areas with low housing elasticity of supply, positive labor
demand shocks tend to increase housing prices with no impact on
employment or GDP.
Recap: the Italian urban system

29. • Italian urban areas have higher employment rate and
higher VA per worker; there are more innovative
activities (KIBS sectors)
• Wage premia are smaller, especially once we take into
account composition effects and housing costs
• Selective migrations
• Possibly the consequences of wage bargaining and
other factors that hamper agglomeration processes
Recap: the Italian urban system

30. Italy
Size
Welfare
High costs, low benefits
(Henderson, 1974)
Other advanced
countries

31. The structure of the project
Urban System Implications for
growth
Historical origins
Agglomeration
benefits
Agglomeration
costs

32. • Lost opportunities for growth due to congestion, low
mobility, and housing rents
• Inequality: among urban centers and within them
(between homeowners and renters)
• Urban dimension of the traditional Italian dualism:
– North too polycentric and congested vs. South with urban centers with
weak agglomeration forces
• Way outs
– Mobilization and utilization of the housing stock
– New innovative programs for public transportations
– Exploit the polycentric model
– The governance of urban centers
Implications for growth and way outs

33. BANCA D’ITALIA
E U R O S I S T E M A