CFE Trento 23 rd February 2018 Regional and Metropolitan data and tools Economic Analysis, Statistics and Multi-level Governance OECD Centre for Entrepreneurship, SMEs, Local Development and Tourism Territorial Analysis and Statistics Unit
Mar 17, 2018
CFE Trento 23rd February 2018
Regional and Metropolitan data
and tools
Economic Analysis, Statistics and Multi-level Governance
OECD Centre for Entrepreneurship, SMEs, Local Development
and Tourism
Territorial Analysis and Statistics Unit
2
1) What are the main databases/indicators that we
collect/produce?
2) What is the scale at which indicators are available?
4) How are the indicators collected and their timeliness in each
database?
3) What are the tools available to visualise/download the data?
5) What are the main challenges in having comparable
indicators at regional level and at city level?
6) What are your major data needs or challenges you
encounter in using the Regional and/or Metropolitan
database?
•
QUESTIONS ADDRESSED
Mandate of the WPTI (2015-2019)
The objective of the Working Party on Territorial Indicators
(WPTI) is to contribute an evidence based analysis of the
regional characteristics, resources, drivers and potential for
development and to improve the understanding of sub-
national patterns and dynamics of structural change in all
types of regions
Databases at subnational level serve the WPTI
Regional Development Policy
Committee (RDPC)
WP on Territorial Policy in Urban Areas
WP on Territorial Policy in Rural Areas
WP on Territorial Indicators (WPTI)
i. Updating, improving, and broadening the OECD
Regional and Metropolitan databases
ii. Broadening the work on identifying and measuring
functional regions
iii. Deepening the work on measuring people’s well-being in
regions and cities
iv. Measuring business demography and entrepreneurship
v. Supporting evidence for policy decisions and evaluation
vi. Engaging in promoting and sharing innovative methods
to integrate geographical and statistical information
(…)
Intermediary objectives of the Working Party
include:
at subnational level:
– Regional database
– Metropolitan database
at country level:
– Subnational Government Structure
and Finance database
– Tourism database
What are the main databases that we
produce?
6
REGIONAL DATABASE
What is the scale at which indicators are
available?
7
National territory
Large regions (TL2)
Small regions (TL3)
Intermediate
Predominantly rural
Predominantly urban
Rural close to a city
Rural remote
Territorial grid Territorial typology
Based on density Based on accessibility
Territorial grid
Regions in each member country have been classified based on two
administrative territorial levels (TLs):
TL2 large regions are defined as the first administrative tier of subnational
government and consists of 398 OCDE large regions. For EU countries TL2
are equivalent to NUTS2, with the exception of Belgium, Germany and the
United Kingdom for which TL2=NUTS1
TL3 small regions are composed of 2 241 small regions , TL3 = NUTS3 for
EU countries
Regions are subject to change over time, especially for EU countries with a
change of classification every 3 years: e.g. as from the 1st of January 2018, data
are submitted following the new NUTS2016 classification, implying split, mergers,
shifts and change in codes. Historical data for the new breakdowns to be sent by
1 January 2020. (http://ec.europa.eu/eurostat/fr/web/nuts/history)
We keep the longest time series when possible
Territorial grid (pdf) is available in the metadata displayed on Dotstat
The most recent changes in the boundaries of
European regions
Country NUTS 2 NUTS 3
Germany
Cochem-Zell recoded from DEB16 into DEB1C and Rhein-
Hunsrück-Kreis recoded from DEB19 into DEB1D due to boundary
change; merge of DE915 Göttingen and DE919 Osterode am Harz
into DE91C Göttingen
Ireland structure revised from 2 into 3 regions regions reassigned and partially relabelled
France many regions reassigned due to revised NUTS 1
structure
Lithuania
Lietuva split into two: Sostinės regionas and Vidurio ir
vakarų Lietuvos regionas (Capital Region and Central
and Western Lithuania Region)
Vilniaus apskritis (NUTS 3 region) reassigned to new NUTS 2
region LT01 Sostinės regionas (Capital Region); recodings due to
establishment of new regions at NUTS 2
Hungary Közép-Magyarország (HU10) split into two: Budapest
(HU11) and Pest (HU12)
NUTS 3 regions of Budapest (previously HU10, now Budapest
HU11) and Pest (previously HU10, now Pest HU12) reassigned
The Netherlands NL121, 122, 123, 338, 339, 322 and 326 recoded into NL124, 125,
126, 33B, 33C, 328 and 329 due to boundary changes
Poland 2 NUTS 2 regions reassigned, 1 new created (capital
region)
several NUTS 3 regions reassigned due to changes at NUTS 1 / 2
level, 1 new created
Finland Kainuu recoded from FI1D4 to FI1D8 and Pohjois-Pohjanmaa
recoded from FI1D6 into FI1D9 due to boundary change
UK
new region Southern Scotland created from parts of
Eastern Scotland and South Western Scotland (now
Western Scotland)
in Scotland, several regions reassigned due to changes at NUTS 2
level; UKN (Northern Ireland) restructured the NUTS 3 level from 5
into 11 regions
Territorial typology
• TL3 regions are defined following 3 categories of typology:
Predominantly Urban , Intermediate, Predominantly Rural
• and 2 sub-categories: rural remote/close to a city
• This typology was initially based upon municipality density. The
European Union updated its typology which relies on the
classification of grid-cells according to pre-established density
and size thresholds
• The OECD has updated its classification of European TL3
regions following these changes. Non-EU countries could also
be updated after formal approval from countries.
STEP 1:
Identification
of urban
clusters
Regional typology
- For OECD non-EU countries, the data source is the
LandScan - High resolution Global Population Dataset
(Census circa 2011)
- For EU countries (including Switwerland,and Norway), the
data source is Eurostat, JRC and European Commision
Directorate-General for regional Policy
Minimum population density threshold:
- For Japan and Korea, 600 inhabitants per km2
- For other countries, 300 inhabitants per km2
Minimum population threshold:
- For Japan and Korea, 10'000 inhabitants
- For other countries, 5'000 inhabitants
If the share of the regional population in urban clusters is:
- higher than 80%, the region is predominantly urban (PU)
- between 80% and 50%, the region is intermediate (IN)
- less or equal to 50%, the region is predominantly rural
(PR)
Note: the typology is defined for small regions (TL3) , except
for Israel (TL2 large regions) STEP 2:
Define
typology - An intermediate region becomes predominantly urban if at
least 25% of its population lives in urban clusters of at least
500 000 inhabitants
- A predominantly rural region becomes intermediate if at
least 25% of its population lives in urban clusters of at least
200 000 inhabitants
- Predominantly Urban (PU)
Regions are classified as: - Intermediate (IN)
- Predominantly Rural (PR)
5. Adjust the
classification to take in
account the presence of
cities (large urban
clusters)
4. Share of urban
population by region:
Attribute a typology to
regions depending on their
share of population in
urban clusters
3. Urban clusters:
Apply a minimum
population threshold to
contiguous densed cells to
identify urban clusters
2. Density threshold:
Apply a minimum density
threshold to 1km2 grid cells
to identify densed
populated cells
1. Data input:
Population grid density of
1km2
Population share by typology
On-going work: refinement of the terminology with distance criteria to split intermediate and rural categories into Remote / Close to a city with the new typology based on grid cells
0
10
20
30
40
50
60
70
80
90
100
Predominantly Urban Intermediate Predominantly Rural
What are the main indicators that we
collect/produce?
6+ datasets on DotStat:
- Demographics
- Economics
- Labour
- Social & Environmental
- Innovation
- Business Demography (as from 2017)
+ Migration (forthcoming)
+ Well-being dataset
+ Regional Income distribution dataset
+ Subnational finance (as from 2014)
Data collection:
- Annual excel questionnaire
- Eurostat and NSO’s web sites
- Cooperation with STI/STD/ENV
- Own tabulation (e.g. Gallup)
Indicators collected:
- Regular pool of indicators but adapted
to your need.
Indicators collected
Resident Population by age and gender
Deaths by age and gender
Number of private households
Inter-regional migration
GDP ; GVA by industry (ISIC rev.4)
Primary Household Income ; Disposable Household Income
Deflators (regional accounts)
Employment ; Labour force ; Young labour force
Unemployment ; Long term unemployment ; Youth unemployement
Employment at place of work by industry
Part-time employment by gender
Labour force attainment by ISCED level
Students enrolment by ISCED level
R and D by sector (expenses and number of personnel)
Percentage of households with broadband access
Rate of young NEET ; Rate of early leavers from education
Number of physicians ; Number of hospital beds
Life expectancy at birth ; Infant mortality ; Transport-related mortality rate
Number of motor vehicles theft ; Number of homicides
Private vehicles
Voters
Municipal waste ; Recycled municipal waste
Air pollution (PM2.5 level) ; CO2 emissions by sector
CO2 emissions by sector
Share of land by type of coverageEN
VIR
ON
ME
NT
INN
OV
AT
ION
ED
UC
AT
ION
SO
CIA
LD
EM
OG
RA
PH
IC
EC
ON
OM
I
CLA
BO
UR
How are the indicators collected and their
timeliness in each database?
Data main update Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun
Excel Questionnaire to
all countries
sent to
countries
received
processed
processed
+ NSO
web
DotStat
GDP/GVA by industry
(eurostat) Dotstat
Employment by
industry (eurostat) Dotstat
Labour LFS (eurostat) Dotstat
Business Demography Dotstat
Patents (STD) Dotstat
Well-being Dotstat
Subnational accounts Dotstat
Income distribution
dataset (every 2 years)
The OECD Regional Databases
4 – Data visualisation
• DotStat: Regional Database http://stats.oecd.org/Index.aspx?DataSetCode=REGION_DEMOGR
How to access DotStat (STATA/Eviews/SAS/Python):
http://oecdshare.oecd.org/itn/kb/as/AS%20Help/DotStatGet_DirectAccess.pdf
• Data Visualisation: Regional explorer http://stats.oecd.org/OECDregionalstatistics/#story=0
Regional well-being: http://www.oecdregionalwellbeing.org
• Regions at a Glance: http://www.oecd.org/regional/oecd-regions-at-a-glance-19990057.htm
Well-being in OECD regions
Province of Trento
AostaValley
Province of Bolzano-Bozen
Sardinia
Liguria
Province of Bolzano-Bozen Province of
Bolzano-Bozen
Province of Bolzano-Bozen
Province of Bolzano-Bozen
AostaValley
Liguria
Lazio
Sicily Calabria Lombardy
Molise
CalabriaCampania
Basilicata
Campania
Calabria
Campania
Sardinia
Safety Jobs Environment Community CivicEngagement
Income Access toservices
Health LifeSatisfaction
Housing Education
Top region Bottom region
Ra
nkin
g o
f O
EC
D r
eg
ion
s(1
to
39
5)
top
20
%b
otto
m 2
0%
mid
dle
60
%
Province of Trento
Provinces
17
METROPOLITAN DATABASE
OECD-EU definition of Functional Urban
Areas (cities)
Why an harmonised definition of cities?
– Policies need to reflect the reality of where people live and
work
– The connections between cities and with surrounding areas
can lead to important changes in how and where economic
production takes place
– Individual cities are interested in comparing their performance
The approach
– It identifies urban areas beyond city boundaries, as integrated
labour market areas
– It identifies urban areas of different size (small urban,
medium-sized urban, metropolitan and large metropolitan)
– It allows comparisons among the different forms that
urbanisation takes
What is the method for FUA?
• The method uses commuting data and population density
calculated for grid spatial units of 1 km ²
• The functional urban areas are defined as densely populated
municipalities (city cores) and adjacent municipalities with high
levels of commuting towards the densely populated urban cores
(commuting zone).
• A minimum threshold for the population size of the functional urban
areas is set at 50^000 population
• It is applied to 30 OECD countries and identifies 1 198 urban areas
For more details on the methodology: “Redefining urban: a new way to
measure metropolitan areas” , OECD Publishing, 2012
A map of Italian FUAs
• In Italy our method
allows us to
identify 74 FUAs
• Total population in
2014 ranges from
52,000 to 4.2
million (Milan)
• 51% of Italian
population live in
FUAs (Milan
represents 13%)
What is the scale at which indicators are
available?
For the metropolitan areas (FUAs > 500,000 inhabitants)
• OECD-EU definition is applied to 30 OECD countries 1 198 FUAs
covering two-thirds of OECD population + 53 in Colombia.
• Shapefiles of the metropolitan areas available on the website:
http://www.oecd.org/gov/regional-
policy/functionalurbanareasbycountry.htm
• List of municipalities by FUA:
http://www.oecd.org/gov/regional-policy/List-municipalities.xls
- Population (level and growth)
- Population density
- Population by age
- Total Area
- Urbanised area (share and
change)
- Polycentricity
- Concentration of population in
core areas
- Sprawl index
- Local units
- Local units in core area
- Territorial fragmentation
- GDP (level and growth)
- Disposable income per
equivalent household
- Income inequality (Gini index)
- Patents application
- Employment (level and change)
- Labour force (level and change)
- Unemployment (level and
change)
- Income segregation (Entropy-
based index)
- Air pollution
- CO2 emissions per capita
- CO2 emissions from transport
and energy sector
Demographic Urban form Territorial organisation
Labour market/Social Environmental Economic and innovation
What are the main indicators that we
collect/produce?
Variable Years available Method
Population (total, core and commuting
zone), population density and by age
2000-2014 Two census data points were collected at municipal level. Inter-census years were interpolated.
GDP (current and constant prices) 2000-2013
Update will be based on the new regional
data
Municipal population data and GDP data at TL3 level were used to downscale GDP at metro
politan level (with the exceptions of Mexico, Canada and Chile where GDP at TL2 level were
used and US for which GDP data at metropolitan level were provided by the Bureau of Economic
Analysis).
Labour (Employment, Unemployment,
Labour force)
2000-2014
Update will be based on the new regional
data
Municipal population data and Labour data at TL3 level were used to downscale Labour data at
metropolitan level (with the exceptions of Portugal, Mexico, Canada and Chile where Labour
data at TL2 level were used and US data for which it was collected from the Bureau of Labour
Statistics).
Labour productivity 2000-2013 Ratio between GDP and total employment in a metropolitan area.
CO2
2005 and 2008
Possibility to compute 2000 data
PM 2.5 estimates 2002, 2005, 2008, 2011 and 2013
Data refer to three-year average
The satellite-based data of air pollution at 1km2 are multiplied by the population living in that
area (using a 1km2 resolution population grid). The exposure to air pollution in a city is given by
the sum of the population weighted values of PM2.5 in the 1km2 grid cells falling within the
boundaries of the city. Finally, the average exposure to PM2.5 concentration is given by dividing
this aggregated value by the total population.
Patents 2000-2008
Possibility to update the data with the
new REGPAT database
Data on patent activity in metropolitan areas are available only for 16 OECD countries .
Extension of countries require correspondence tables between zip –municipalities FUAs.
Housing 5 countries: Canada (2006), US (2012),
Norway (2001), Chile (2002) and Mexico
(2010)
New data would need to be collected for other OECD countries before including it in the
Metropolitan database.
Urban land 2000 and 2010 The finest Global Land Cover dataset: resolution 30m2.
Income
How are the indicators collected and what
are their timeliness?
• OECD.Stat http://stats.oecd.org/Index.aspx?Datasetcode=CITIES
• http://measuringurban.oecd.org/
What are the tools available to
visualize/download the data?
• To improve our capacity to provide useful
information for Trento
Thank you!
Feedback?