Citybench - Frequently Asked Questions

Every time I access the CityBench Webtool, the city I am in (or near to) is initially selected. Why is that?
The CityBench Webtool is able to derive your (approximate) location from the IP address of the device you are using. We assume that this is also the city you are most interested in. In case CityBench is not able to determine your location, the city of Luxembourg will be initially selected.

What is a Larger Urban Zone (LUZ)?
Although called "CityBench", this Webtool is aimed at benchmarking and comparing Larger Urban Zones (LUZ) rather than cities. A LUZ is "…an approximation of the functional urban area extending beyond the core city"; the core city being "…the city as defined by its administrative and/or political boundaries" (definitions from Eurostat). In practice this means that a LUZ is virtually always larger than the city it contains and that values collected for LUZ level will, to a greater or lesser extent, differ from (core) city values.
It should be noted that LUZ delineations are not fixed, but tend to be revised on an irregular basis. The latest version of the LUZ nomenclature is LUZ 2012, which in some cases differs quite considerably from the previous version (LUZ 2004). To prevent inconsistencies in the indicator values, both between different LUZ and between different years, CityBench only uses LUZ data calculated for or derived from the 2012 delineations. For some indicators this might mean that time series are not, or to a limited degree, available.

What is NUTS3?
According to Eurostat, "The NUTS classification (Nomenclature of territorial units for statistics) is a hierarchical system for dividing up the economic territory of the EU …". The most commonly used levels are:
- NUTS1: major socio-economic regions (not used in Webtool)
- NUTS2: basic regions for the application of regional policies (not used in Webtool)
- NUTS3: small regions for specific diagnoses (several indicators in Webtool)
In case of the NUTS3 indicators in the Webtool, the NUTS3 values have been aggregated to the metropolitan region (MR) level, as an approximation of the LUZ level. This was done using an MR - NUTS 3 (2006) correspondence table (not available online). Depending on the indicator, NUTS3 values were either simply aggregated or, more laboriously, assigned weighting factors, most notably population number or area (again, depending on the indicator). Indicators derived from NUTS3 data are marked as such in the Webtool.

What is a metropolitan region (MR) and how does it compare to a LUZ?
According to Eurostat, "Metropolitan regions are NUTS3 regions or a combination of NUTS3 regions which represent all agglomerations of at least 250 000 inhabitants". An MR is considered the NUTS3 approximation of a larger urban zone. Please note that although many MR - LUZ pairs show a good correspondence, correspondence is poor in some cases (MR much larger than LUZ or vice versa). Indicators collected at MR level are marked as such in the Webtool.

Which LUZ are included and why?
171 LUZ are currently included in the Webtool. They were selected according to the following criteria:
1) Inclusion of all European cities that are part of another online city comparison tool, the OECD Metropolitan Explorer.
2) Inclusion of capitals of European countries not included in OECD to cover ESPON Space (i.e. EU27 plus Iceland, Liechtenstein, Norway and Switzerland).
3) Inclusion of all remaining LUZ with a population number of > 400,000.
4) Inclusion of additional LUZ in under-represented countries, i.e. countries with only one or even zero LUZ with a population of > 400,000 by including a second city, provided that its population exceeds 200,000.
Please note that for Liechtenstein / Vaduz no LUZ, and for both Liechtenstein / Vaduz and Reykjavik no MR have been defined by Eurostat. Therefore, where appropriate NUTS3 values for Liechtenstein and Reykjavik have been included as an approximation of the MR level values.

Why isn't my city included in the Webtool?
If you are in a city which does not comply to any of the criteria specified above, it is currently not included. However, as indicator data availability increases, more and more (smaller) LUZ will be added to the Webtool, possibly including your city as well.

Where can I find an overview of included cities and indicators?
Please click here to download an overview of cities and indicators included in the Citybench Webtool.

What do the various indicator themes represent?
The indicators are grouped according to 'themes', each covering a different topic.
- Theme Economy and Population: provides information about population numbers, GDP, unemployment, etc.
- Theme Connectivity: provides information on the degree to which a LUZ is connected to 'the rest of the world'.
- Theme Demography: provides information on the build-up of the LUZ population.
- Theme Social Media: includes indicators derived from social media data harvesting.
- Theme Investment Climate: provides indicators potentially of interest to investors.
- Theme Environment: provides an indication of the environmental / air quality of a LUZ.
- Theme Smartness: provides an indication of the degree to which a LUZ is prepared for future developments.
- Theme Quality of Life: provides an insight into the performance of a city as perceived by its inhabitants.

How are the social media indicators derived?

How can I get more information for a specific indicator?
Below each indicator is an 'info' link. When clicked a popup window shows additional information about the indicator, such as a description and the data source URL.

Why are some indicators not available for the city or cities I'm interested in?
Because of gaps in the source data, unfortunately completeness for most indicators is not 100%. This means that they do not cover all LUZ included in the Webtool. A small circle to the right of each indicator name shows its degree of completeness. The dark blue portion is the proportion of LUZ for which indicator values are available.

Is it possible to show indicator values for different years (time series)?
Yes, if available, time series for an indicator are displayed in the graphs / charts when comparing LUZ. Please note that when viewing LUZ similarity in the map or radial view, by default the most recent year for which indicator data was collected is displayed. However, if other years are available, they are shown under the indicator name and can be selected.

How can I compare two or more cities?
The CityBench Webtool allows the comparison of up to four LUZ, based on up to three indicators. To select LUZ for comparison, just click on the buttons below the map of Europe: 'Select the 1st (2nd/3rd/4th) city'. Also, select one, two or three indicators on which the comparison will be based by clicking 'remove' and/or 'Select a 1st/2nd/3rd indicator'. Once you have selected one or more LUZ and one or more indicators, click 'Compare'.

Rather than comparing between cities, can I compare a city to other geographic units?
Yes, you can. The CityBench Webtool has the option to compare LUZ indicator values to the average value of the country it belongs to or to the European average.

How is the similarity between cities calculated?
By similarity we mean the degree to which indicator(s) values are comparable.
Two processing steps are involved in deriving similarity values:
1) All indicators are normalized, i.e. for each indicator the lowest LUZ value is assigned a value of 0 and the highest a value of 1.
2) For one, two or three normalized indicators, the 'Euclidean' distance between the indicator(s) values for one LUZ and one or more other LUZ (or country / European average) is calculated. Euclidean distance (or Euclidean metric) is the distance between two points as measured with a ruler and using the Pythagorean formula to derive the (metric) distance. The more indicators are selected, the more dimensions are involved in the calculation (one dimension for each indicator added).
For example, if you selected 3 (normalized) indicators (a, b and c) and two LUZ (A and B) then the similarity between these LUZ is calculated as follows:
sqrt ((aA - aB)2 + (bA - bB)2 + (cA - cB)2).
If the calculated distance is 0, there is full similarity; if the distance is equal to the sqrt(number of dimensions), there is maximal dissimilarity. In other words, the closer the number is to 0, the higher the degree of similarity between the selected LUZ, based on the chosen indicators. Conversely, the closer the number is to 1 (or even exceeds 1, which occurs as more indicators are added), the more different the LUZ are.

What is shown on the different tabs: Map / Radial?
Being the main tab, the Map view shows how all included LUZ compare based on one, two or three indicators. The LUZ currently selected serves as the reference LUZ, to which the others are compared. The size of the circle representing a LUZ corresponds to the similarity of each LUZ to the reference LUZ: the larger the circle, the greater the similarity (and vice versa). See for an explanation of the calculation of similarity: How is the similarity between cities calculated?

The Radial view provides, for one, two or three indicators, an alternative way to show the similarity between the target LUZ and all other or selected other LUZ. The radial view consists of a two-dimensional space with the target city in the center and the other / selected cities circling around it. The LUZ are positioned more or less according to their actual geographic location. The less similar a LUZ is to the target LUZ, the further away from the center it is placed. The LUZ circle sizes in the radial view represent the actual values of the indicator selected on the right. See for an explanation of the calculation of similarity: How is the similarity between cities calculated?