10/2007: Hervé Carré


  Hervé Carré
Director-General of Eurostat
(European Commission, Brussels)


‘Towards a common framework for the collection and harmonisation
of migration and asylum statistics in the European Union’


Ü Eurasylum: After two years of discussions within the EU institutions, the Proposal for a Regulation of the European Parliament and the Council on Community statistics on migration and international protection was adopted in July 2007. The Regulation aims to establish a common framework for the collection and compilation of Community statistics on international migration, as well as to reduce the impact of definitional and data source differences on the comparability of the statistics. The statistics to be collected under the new Regulation will, as far as possible, follow the United Nations Recommendations on Statistics of International Migration, and they will be broadly based on Eurostat’s existing annual migration data collection system, and the monthly data collections on asylum and enforcement measures against illegal migration. An additional migration-related topic to be covered will be the legal immigration of non-EU citizens. Can you guide us through the main features and implications of this new Regulation?

Ü Hervé Carré: The adoption of the Regulation 862/2007 in the summer of 2007 was a major step towards the provision of reliable and harmonised statistics on migration and asylum in Europe. Most important areas of Community statistics have a clear basis in European law, defining the responsibilities of Member States and of Eurostat in terms of the collection, transmission and publication of data. The migration statistics domain had been unusual in not having a legal base, being instead governed by a series of voluntary agreements between Eurostat and the data suppliers in Member States. While this may have been appropriate in the past, it became clear that the growing policy importance of this subject at both national and European level meant that a more formal approach was necessary.

It is worth mentioning that although Eurostat performs many of the functions that might be expected of a national statistical institute, Eurostat is not involved in the primary collection of data. The data used for European statistics are collected by the national statistical institutes and other government agencies, before being compiled in a harmonised manner by Eurostat. A complication in this domain is that the data are supplied by a wide range of institutions and agencies at national level  including the national statistical institutes, Ministries of Justice and Interior, Immigration and Border Control Services, and Police authorities. One of the most difficult aspects for Eurostat is maintaining good contacts with all of these different data suppliers.

The finalisation of the Regulation also represented the end of four years of consultations and negotiations with data suppliers and statistics users. Early on in this process, it was decided that the close links between the statistics and the rapidly developing EU policies and legislation in this domain meant that the Regulation should be jointly developed and proposed with our colleagues in the Commissions Directorate General for Justice, Freedom and Security. The first drafts of the Regulation were actually prepared back in the autumn of 2003. There was a great deal of detailed technical discussion with Member States, international organisations and academic experts  and much redrafting  before a formal Commission proposal for the Regulation was sent to the European Parliament and the Council in 2005.

An unexpected benefit of this consultation process was that Eurostat, the data suppliers and the users of the statistics were forced to reassess what types of data collection were possible, and what statistical outputs were really needed. This was of particular value in this domain where a number of statistical activities had developed in a largely ad hoc manner to keep pace with very rapidly changing policy needs following the 1999 Treaty of Amsterdam that gave the Commission policy competence in immigration and asylum matters.

The Regulation covers most of Eurostat’s existing statistics on migration related issues. Statistics on immigration and emigration flows, together with statistics on the citizenship and country of birth composition of the resident population, provide information on the impact of migration on the size and structure of the population. Statistics on asylum applications and the subsequent decisions to grant or refuse refugee status or other types of international protection will be adapted somewhat under the Regulation to better meet user needs for these data and taking into account the burden on data providers. For example, asylum applications statistics will be collected on a monthly basis as these are needed to allow a continuous monitoring of short-term variations in the origin and numbers of asylum seekers. In comparison, data on appeals against asylum decisions are relatively complex to collect and are not needed so frequently  and so will only be collected annually. For the statistics on measures against unauthorised migration  refusals of entry at the frontier, persons found illegally present, and removals  the introduction of the Regulation will see a move away from rapid collection of monthly statistics, to the annual collection of more reliable reference data. We are also expanding the publication of these statistics to make sure they are equally accessible for all users.

The only new area of statistics covered by the Regulation is that of residence permits issued to non-EU citizens. These statistics offer a useful insight into the reasons for immigration  as a distinction can be made between permits issued under different immigration rules regarding the reunification of families, labour migrants, and persons admitted as students. Much of the other data available on migration flows and population stocks of migrants  such as data from population registers – does not give information on the reasons for migration or the migrants activities in the destination country.

A further aspect of the Regulation is that most of the statistics to be collected will include a disaggregation by age and sex. Some statistics collected in the past  particularly asylum and control of immigration data – often lacked this information. It is however important that these disaggregations be made available wherever possible. For example, this information is required as part of the monitoring of policies aimed at preventing the trafficking of women and children.

The challenge now will be to implement the legislation in an effective manner through the adoption of implementing Regulations in 2008. Many details remain to be defined  such as the formats of tables and arrangements for transmitting the data to Eurostat. This is a collective exercise in which we will continue to work closely with the national data suppliers and users of the statistics.

Ü Eurasylum: Whilst the new Regulation requires that Member States produce statistics that meet as closely as possible harmonised definitions, it does not place a duty on the Member States to introduce completely new data sources or to compel changes to the administrative systems for immigration or asylum. However, Member States will have to explain their choice of data sources and explain the anticipated effects of their collection methods on the degree of compliance with the harmonised definitions. This information will be used as a guide to the interpretation of the statistics. Can you comment, briefly, on some of the current discrepancies between national data collection methods within the EU, and the anticipated challenges in achieving a common and harmonised framework for the collection of migration and asylum statistics in the European Union?

Ü Hervé Carré: A central aim of the Regulation was to overcome the current problems of missing data and poor harmonisation. Currently, differences in the definitions used mean that comparisons between countries can be difficult or misleading. Member States differ greatly in terms of the data sources used for migration statistics and the statistical definitions that are applied. Whereas some Member States base their migration flow and migrant population stock statistics on population registers, others may use sample surveys or data extracted from administrative procedures such as the issuing of residence permits. Statistical definitions have generally followed closely the national administrative rules and data sources. Therefore, if a person needs to register formally as a resident for a stay of three months or more, then this three-month duration of stay is used in the statistics to distinguish between a visitor and a migrant. Other countries have applied different time limits in their definitions or have applied additional or alternative conditions  relating for example to the reason for entry or the rental or purchase of a dwelling.

To ensure comparability, the data supplied by Member States must be based on harmonised statistical definitions. The definitions in the new Regulation are themselves based on the recommendations prepared by the United Nations, and on current and forthcoming EU legislation on asylum and immigration. However, although data must meet harmonised definitions, because of the great differences between Member States in terms of data sources used, it was not feasible to legislate for a particular data source to be used. Member States remain free to use any appropriate data source according to national availability and practice. Therefore, the Regulation can be seen as a measure that will harmonise statistical outputs rather than the data inputs to the statistics.

The new Regulation also allows scientifically valid estimation techniques to be used where, for example, directly observed data are not available or where data based on national definitions must be modelled to fit the harmonised definitions. Eurostat is funding a project headed by the Netherlands Interdisciplinary Demographic Institute (NIDI) and involving migration statistics experts from several Member States that is developing advanced estimation techniques for migration statistics and which will provide assistance to Member States in applying these techniques.

Certain areas of migration statistics present particular problems in terms of the availability of data. Measures designed to facilitate travel across the EU internal borders and to make it easier for EU citizens to live and work in another Member State have had the unfortunate effect of reducing the availability of administrative data. With reduced border controls and fewer requirements to apply for permission to stay, there are fewer opportunities for statistical data to be derived. Similarly, emigration is extremely difficult to measure as departures are far less well recorded than arrivals in registers and other administrative systems. Emigration and the migration of EU citizens are both areas that will require attention under the estimation project just mentioned. In addition, developments will be encouraged in the use of surveys to supplement existing migration data sources and in the exchange of data between national systems.

The harmonisation of residence permit statistics is made difficult by the complexity of the immigration rules in many Member States. Although the broad categories of labour migration, student and family permits are found everywhere, there are many differences in terms of the sub-categories of permit, the conditions for obtaining such a permit, and the rights that such a permit holder is granted. One example of this in an area with high policy interest concerns special immigration schemes for highly qualified workers. Although such schemes exist in a number of Member States, there are differences in the definition of ‘highly qualified’. The European Commission is preparing proposals designed to harmonise more closely policies and procedures for highly qualified immigration. The residence permit statistics will need to be developed and adapted to reflect policy needs in this important area.

Ü Eurasylum: According to one of the conclusions of the EU-supported THESIM project (Towards Harmonised European Statistics on International Migration), the harmonisation of data on enforcement measures against illegal immigration might be more difficult to achieve across the EU, given that the reliability and comparability of statistics on illegal migration are difficult to establish. In particular, according to THESIM, data on refusals of entry to any of the EU Member States should at least provide the reasons for such refusals, while data on apprehensions of illegal migrants should be disaggregated at least according to reasons for apprehension, the location of apprehensions (border or in-country), whether they involve nationals or non-nationals (e.g. for trafficking purposes), the directions of movements (apprehension upon entering or leaving the territory), and the countries of origin or destination of movements. Data on removed aliens should further qualify the categories of removed aliens and types of removal (e.g. expulsion, deportation order, monitored departure, etc.). Can you discuss some of the key challenges and proposed solutions related to the harmonisation of statistics on illegal migration across the EU?

Ü Hervé Carré: First, it is important to note that the data collected by Eurostat relate to measures against illegal migration and stay. There is no attempt to estimate the flow or population stock of unauthorised migrants.

As with other aspects of the migration statistics legislation, the statistical content has been defined in very close reference to the current and foreseeable policy needs. These needs for statistics must always be balanced by what information it is feasible and cost-effective to collect. However, many of the disaggregations proposed by THESIM are included. For example, the Regulation states that statistics on refusals at the frontier are to be disaggregated in accordance with the Schengen Borders Code  which includes reason for refusal. This link to legislation concerning immigration and border controls ensures the relevance of the statistics produced, and also reduces the burden on data suppliers as the statistics can be based directly on the outputs of administrative systems. Similarly, the Regulation also foresees future disaggregations of the statistics on apprehended illegal migrants by reason for and place of apprehension.

There will be difficulties in ensuring the harmonisation of these statistics  in common with all areas of migration statistics. However, one advantage with the statistics on measures against illegal migration is that there is considerable commonality of data sources – all Member States use administrative data derived from the enforcement procedures. Also in common with other areas of migration statistics, the ongoing progress towards more harmonised policies means that the concepts and processes which underlie the statistics also become more comparable.

There generally appears to be broad agreement now among the agencies responsible for statistical data on migration and asylum that positive efforts must continue to ensure that these statistics can better meet the needs of all users. With the new Regulation, and with a positive approach from all concerned at national and European level, we can be optimistic that both the availability and comparability of European migration statistics will improve greatly over the coming years. My optimism in this regard has been confirmed at last month’s conference of the Directors General of National Statistical Institutes in Budapest. The European Statistical System is committed to develop an effective framework of migration and asylum statistics. Never before have policy developments in the Justice, Freedom & Security fields been co-ordinated so closely with the adoption of statistical legislation and strategic thinking in the social statistics area.