/ EUROPEAN COMMISSION
EUROSTAT
Directorate E: Sectoral and regional statistics
Unit E-4: Regional statistics and geographical information

Summary of the final meeting on the Labour Market Areas (LMAs) 2016/2017 grants

28 June, Luxembourg

  1. Approval of the agenda

The chair of the meeting, Teodora Brandmuller(Eurostat, Regional Statistics and Geographical Information unit) welcomed all participants. All participants briefly presented themselves. The agenda was approved.

  1. Nature of the meeting

All National Statistical Institutes (NSIs) participating in the project were invited to the meeting as well as stakeholders from other units within Eurostat and from other European Institutions. All NSIs participating in the project were present except for Statistics Finland. See list of participants in Annex 1.

  1. List of points discussed

Item 2: Presentations of NSIs on the project results

Presentations

The presentations given by the representatives of the NSIs participating in the project were structured in a way to demonstrate the main achievements according to the grant program and to answer to the following questions[1]:

•How do you assess the overall project experience?

•What are the lessons learned?

•What were the main challenges faced?

•Would you propose any changes to some specific details of the method or terminology used etc.?

•What are your post-project recommendations and your future plans regarding the LMAs?

•Do you have further comments to the Eurostat's final report on the LMAs project?

Project's overall assessment

The participants in the final meeting expressed a common satisfaction of the project results: national and cross-border LMAs have been created; guidelines on the workflow, scientific papers, experimental data at level of LMAshave been produced, a lot of co-operation and motivation for work have been observed.

The R package has been significantly improved compared to the previous version. It is free of charge and publicly available.

Various technical and philosophical questions have been discussed at several forums as part of the knowledge exchange and building of sustainable capacity.

Lessons learned

Differences in the commuting patterns, in the urban systems and the size of the building blocks can have a strong influence on the outcomes at country level (the experience of Portugal and Poland). The Statistical Office of Hungary developed a systematic and automatized way of testing the parameters, which is seen as a good basis for analysing the outcomes and the method itself. Furthermore, the analysis of border stability based on different parameters is a good basis for evaluation of the outcomes.

The SMART measure calculating the commuting links that is behind the R package led to underestimation in the case of Hungary. That is why an alternative measure (INTRAMAX) has been used.

Registered data, even with higher geographical resolution, not always gives better results than Census data and should be used cautiously (the experience of Portugal).

All countries that have time series confirmed that the number of LMAs decreases over time.

Main challenges

  • The decision on the optimal set of parameters has been recognised by the participants as the main challenge. The countries had to work with vague recommendations to support the choice of the parameters. The output is very sensitive to the parameters. That is why countries with similar numbers of LAUs and NUTSs obtained different number of LMAs of different sizes.
  • Several countries found the parameters' names unintuitive (Netherlands and Poland).
  • In some specific cases such as building block composed by urban and rural part, the fine-tuning still has to be done manually.
  • Countries working with registered data faced the problem of the identification of the actual place of work.
  • With respect to the delineation of cross-border LMAs, the biggest challenge was to estimate the missing component of the commuting matrix followed by the treatment of different coding systems used in different countries.

Proposals for improvement of the method

  • The definition of common evaluating criteria for selection of the optimal set of parameters is seen as crucial for the harmonisation at EU level.
  • Several countries recommended more appropriate names for the parameters that would fit better to their real functions.
  • The participants agreed that the introduction of common commuting distance threshold is needed in order to eliminate fake commuting links between remote LAUs.
  • The spatial contiguity requirementcould be defined more precisely.

Post-project recommendations and future plans

The NSIs have the following recommendations and future plans:

  • Eurostat and the Member States need a common view on the scale (large LMAs comparable with Metropolitan FUA – Eurostat and OECD concept – or LMAs to breakdown Metropolitan FUAs).
  • Clearer recommendations on the parameter selection
  • More guidance on sensitivity analysis
  • Further improvements of the R package;
  • Defining labour market areas on the basis of the Population Census 2021 data; update between the Censuses;
  • Testing new types of data: samples/administrative/big data/integrated data and looking at new emerging types of jobs;
  • Study of the LMAs' evolution over the time;
  • Dissemination of data at LMAs by occupational categories, gender, age groups, earning groups, mode of travel to work, NACE rev. 2 economic activities etc. including using Small Area Estimations; attributing some regional statistics to LMAs with EU relevance;
  • Comparison of the LMAs with the spatial profile of the transport infrastructure;
  • As cross-border data are missing a methodology for imputation using auxiliary information should be developed;
  • To facilitate delineation of cross-border LMAs, EU harmonised data formats, including coding system will be needed in the future;
  • Study on ways to adapt small LMAs to EU policy needs;
  • Further trainings to third countries (without experience with LMAs);
  • Further promotion of the concept to the national stakeholders;
  • Publications devoted to LMAs;
  • Development of visualisation tools;

Item 3: Presentation of Eurostat on the future of LMAs

Presentation

Complementing the ideas given by the NSIs and inspired by the results of preliminary stakeholders' consultation with DG REGIO, DG EMPL, DG MOVE, DG MARE, OECD and ESPON─ from Eurostat's perspective the following future actions could be envisaged:

  • Use of LMAs to optimise the samples of the social surveys;
  • Definition of urban-rural LMAs to study the urban-rural partnership, small and medium urban areas and major economic conurbations;
  • Design of alternative LMAs to discover commuting patterns, education mobility and spatial skill's mismatches; Maritime industrial districts;
  • Examination of over- and under-bounded LMAs in order to improve sub-national employment and GDP data distorted by administrative boundaries and to assess the magnitude of the distortion;
  • LMAs as Housing Market Areas.

Discussion

The discussion on item 2 and 3 was structured as a SWOT analysis. Participants were asked to write down short comments/statements and afterwards to vote for the most relevant comment(s). The results are as follows:

Strengths

  • The model requires employment and commuting data only;
  • Open source and methodology (R package and documentation);
  • A way to create functional geography based on sound theory;
  • Useful analytical tool;
  • Stable expert group of countries.

Weaknesses

  • The targeted scale shall be recommended by Eurostat and agreed by the MSs
  • Policy makers at country level need more time to consider the LMAs as a geography for taking decisions;
  • The purpose is not clear enough;
  • The output is too sensitive to the parameters’ selection;
  • Different building blocks are used.

Opportunities

  • Development of dissemination tools would increase the visibility;
  • To create better guidelines on parameter selection and clear criteria;
  • Harmonisation and comparability;
  • Two levels of LMAs;
  • Policy relevance could be further recognised;
  • LMAs as a tool rather than a new territorial classification;
  • Big data, administrative data and Small Area Estimation;
  • Industrial districts as alternative LMAs;
  • Geography for place-based policy;
  • LMAs as 'basins of life' - opportunity to analyse living, working, housing and commuting in a consistent system;
  • EU-harmonised data formats and coding system;
  • Cross-border LMAs all over Europe.

Threats

  • The medium- and long-term strategiesare not identified yet;
  • Due to its complexity the method can be unattractive for the policy-makers;
  • Additional burden for the NSIs;
  • Exclusion of cross-border data results in different national LMAs;
  • Different input data sources;
  • Still different definitions across countries;
  • Too many (small) LMAs or significant deviation between countries.

Open questions

  • Do we want to have LMAs as a fixed geography (a new nomenclature in TERCET?) or as atoolkit to compute LMAs and analyse results that can be used according to the user needs?
  • How to keep balance between comparability at European level and applicability to local, regional, national and European policy needs?
  • Do we need two levels of LMAs?
  • Which administration would implement policy based on LMAs' analysis?
  1. Conclusions/recommendations/

On behalf of Eurostat, Teodora Brandmüller concluded that the project was a great success: since the work done by the participants went beyond the tasks included in the grants' technical specifications and beyond the expectations of Eurostat. She thanked all participants for their outstanding work throughout the project. Colleagues from ISTAT in particular have demonstrated inspiring dedication to the project which has led to excellent results.

All participants agreed that this group of expertsshould be kept also in the future by continuous exchange of LMAs related information, which will further increase the knowledge and improve the advances. Eurostat is expected to identify the short- and mid-term goals towards harmonised LMAs.

  1. Next steps

Eurostat will make an assessment of the output by looking at several examples. Based on the results of this assessment,the reports produced, the discussions during this meeting, and a foreseen consultation with the methodological unit, Eurostat will prepare a position paper regarding the future of the LMAs. This paper will be circulated to the grant recipients and will be present it to the Working Group in October 2017.

ANNEX: List of the participants

1 / JOHAN VAN DER VALK – CBS NETHERLANDS
2 / JULIA SCHMITT – CBS NETHERLANDS
3 / KEZÁN ANDRÁS – HCSO HUNGARY
4 / ELSA SOARES - INE PORTUGAL
5 / FRANCISCO VALA - INE PORTUGAL
6 / RENI PETKOVA - NSI BULGARIA
7 / LUISA FRANCONI - ISTAT ITALY
8 / SANDRO CRUCIANI - ISTAT ITALY
9 / DOMINIK ŚLIWICKI - POLAND
10 / PAWEŁ STOPIŃSKI - POLAND
11 / MACIEJ RYCZKOWSKI - POLAND
12 / TEODORA BRANDMÜLLER - EUROSTAT
13 / OLIVER MÜLLER - EUROSTAT
14 / VALERIYA ANGELOVA - EUROSTAT

1

[1] The presentations are available on the cros-portal