Local level database for socio-ecomomic indicators

Français non disponible
English

This output is a database of estimated local level data for indicators of socio-economic inequalities modelled through the disaggregation of NUTS2/NUTS3 level data using spatial data estimation techniques.

Existing statistical sources of data for indicators that measure territorial inequalities, such as, income, poverty, education and other socio-economic indicators are currently available at NUTS 2 level. The project IMAJINE suggests that in order to address territorial inequalities, these topics should be analysed at the local level (spatial distribution of inequality). To address this issue, one of the key aims of IMAJINE is to provide an inclusive and homogenous database at the local level for several EU countries. The database can help look at the ‘winning driver’, which refers to the variable which is most important in explaining the inequalities at a local level.

The project performed numerous literature reviews as a first step for the setup of the local level database.

Literature review

First, a literature review was conducted to assess the existing official data and socio-economic indicators, demonstrating the importance of the disaggregated analysis to explain territorial inequality trends and inform place-based policy initiatives. Second, a literature review on disaggregation methodologies of spatial data, notably, areal techniques and micro data-based techniques, and their weaknesses was conducted to justify the use of the Generalised Cross Entropy (GCE) method. Third, a literature review of the territorial inequalities in the EU to show that although there are many studies at the regional level an analysis at the local level was lacking.

Following these extensive literature reviews, researchers established two indicators that are essential for the study of territorial inequalities and which, until now, simply did not exist at the local level for many EU countries:

Average household income

Using the indicator of ‘household income’ is the common procedure when analysing income inequality. Income carries a large weight in the level of wellbeing and life satisfaction both at individual and also at the territorial level (either by locality or region), and is therefore considered a good indicator of economic welfare and development. However, when it comes to the spatial level, due to confidentiality reasons, in Europe in the best of the cases only information at the NUTS 2 region level of residence is provided. This implies that any study of European income inequality is spatially restricted to large, heterogeneous and, in many cases, geographically extensive regions. The broad NUTS 2 administrative regions do not capture the rural-urban divergence or core-periphery dynamics that might be present within the regions, nor do they allow the analysis of local development using this indicator.

Population at risk of poverty or social exclusion

The poverty indicator used is AROPE, a multidimensional single indicator, which measures the rate of people at risk of poverty or social exclusion.

The AROPE indicator considers that an individual is at risk of poverty or social exclusion if he/she meets at least one of the following three criteria:

Lives in a household with an income (including social transfers) below the poverty line, which is defined as an income that is 60% of the median of the national income’s equivalent in consumption units.
Lives in a household where its members cannot afford at least four of the nine basic consumption needs defined for Europe.
Lives in a household with low work intensity, defined as the ratio between the number of months actually worked by all the members of the household and the maximum number of months that all people of working age in the household could theoretically work.

Like the income indicator, AROPE is very limited in its spatial dimension as it is only available for NUTS 2 regions.

Some additional indicators were also collected when available at the local level.

Additional indicators

Some additional local variables, related to the level of education, dependency or immigration of the localities, were collected for the EU countries where that information was available at such a level of spatial disaggregation. This completes the picture so that EU territorial inequalities in terms of income, poverty, education, immigration or ageing are given a new spatial perspective. Not only can spatial disparities between EU regions be analysed, but also, the territorial inequalities within the regions.

The project collects variables at the local level, where they exist, with the help of the corresponding national institutes of statistics (mainly census household data) of the different Member States and econometrically estimates figures at the local level (LAU 2 i.e. municipalities, districts) when such information exists at regional (NUTS 2 or 3) level, but is not available at the local level (e.g. Germany provides census data only at regional/lander level). The local level database covers the following countries: Belgium, Denmark, Finland, Germany, Greece, Ireland, Italy, the Netherlands, Poland, Portugal, Romania, Spain, Sweden, the UK.

In order to offer local estimates for several EU Member States which are consistent with official databases (EU-SILC) IMAJINE applies the Generalised Cross Entropy (GCE) method to solve the problem of spatial disaggregation.

Generalised Cross Entropy (GCE) method

This methodology is based on the framework of Maximum Entropy. In this framework, the variable of interest has a probability distribution with an unknown probability for each value. The basic idea of this type of methodology is to obtain the estimation with the highest degree of uncertainty that at the same time is able to fulfil the conditions from observable data. The Generalised Maximum Entropy (GME) estimator has been applied to the estimation of linear regression equations. The GME estimator will choose the distribution of probabilities that deviate least from a situation of maximum uncertainty as the optimal solution.

By applying this novel methodology, it is possible to obtain consistent estimates of average household income at the local level and by natural intervals.

This methodology was tested in four study countries (the UK, France, Spain and Portugal) by disaggregating the two indicators at the local level. Detailed maps by study country have also been produced. The assessment of the reliability of local estimates proved that the proposed methodology for estimating data at the local level provides statistically satisfactory socio-economic indicators.

Review of official data

Literature review on disaggregation methodologies

Report on inequality indices at local level

Dataset of Local Indicators of Territorial Inequalities

Spatial Disaggregation of Social Indicators: An Info-Metrics Approach Esteban Fernandez-Vazquez, Alberto Diaz Dapena, Fernando Rubiera-Morollon & Ana Viñuela Social Indicators Research volume 152, pages 809–821 (2020)

Presentation at the Good Practice Workshop of the Evaluation Helpdesk

Relevance for monitoring and evaluation of the CAP

It is important to have data at local level to develop and evaluate local-based policies. Without this, the only way to understand what is really happening is through case studies, but it is impossible to cover all regions in all countries via case studies. The local level database for socio-economic indicators is therefore very relevant for the evaluation of rural development policy.

First, for identifying baseline indicators for evaluation, especially for evaluations of local development strategies in the context of LEADER.

Second, for the assessment of the effects on local development through measures implemented using the LEADER approach. The evaluation of the effects of rural development policy on local development has been constrained by the limited availability of data at the local level.

Third, for impact evaluation, specifically for assessing geographical impacts relative to local patterns of inequality. The assessment of socio-economic impacts of the current programming period used impact indicators that deal with standard socio-economic variables such as income and poverty.

The choice of the unit of analysis for impact evaluation, depends on the evaluation approach adopted. One of the recommended approaches for impact assessment is the Propensity Score Matching (PSM) which enables the appraisal of the counterfactual and therefore the assessment of net impacts. This approach needs data at the lower spatial level of LAU 2, but often data is not available at this level and the NUTS 3 level is used instead as a secondary option (for more information on this, see the Helpdesk Guidelines on Assessing RDP Achievements and Impacts in 2019).

Therefore, the IMAGINE local level database can be a useful source for local level socio-economic variables for the countries it covers and can be used by them without any particular adaptations, except for adding more data if required. It offers a new dataset with the possibility to use in many applications, including for analysing socio-economic trends at a more highly disaggregated level than currently available. The same approach can also be used for collecting local level data in other Member States.

The local level database is transferable to other Member States (outside the study countries) provided there is cooperation with national institutes of statistics as well as the involvement of experts with knowledge of econometrics in order to use the GCE method and transform any regional level data into local one.

Last modification date:

09/12/2021

Project details

IMAJINE:
Integrative Mechanisms for Addressing Spatial Justice and Territorial Inequalities in Europe

Objectives

Formulate new integrative policy mechanisms to enable European, national and regional government agencies to more effectively address territorial inequalities within the European Union.
Imagine a future for European regions in which the distribution of resources is consistent with principles of social and spatial justice.
Respond to a pressing need to re-appraise the appropriateness and efficacy of existing policy instruments for tackling territorial inequalities, and to consider and develop alternative mechanisms.
Produce new primary data, apply new analytical tests to secondary data and integrate the results with insights from relational geographical theory and the concept of spatial justice.

Timeframe:

2017 to 2022

Contacts of project holder:

Professor Michael Woods, Aberystwyth University (UK)

Project website(s):

IMAJINE

CORDIS database

Territorial coverage:

Belgique, Finlande, France, Allemagne, Grèce, Irlande, Pologne, Roumanie, Espagne, Les Pays-Bas

Relevant CAP Objectives:

8. Vibrant rural areas

Data Collection Systems Used:

Population census
Eurostat

Associated Evaluation Approaches:

Impact evaluation ex post
Impact evaluation ongoing

Type of Output:

Database/ data registry

Spatial Scale:

Sub-regional / local
Regional

Keywords:

impact assessment
income
LEADER/CLLD
local development
poverty
socio-economic effects