Final Report Summary - SOM (Support and opposition to migration - A cross national comparison of the politicisation of migration)
Executive summary:
Since the 1960s, substantial numbers of immigrants migrated to Western Europe. At least from the mid-Nineties onwards, the presence and integration of immigrants became a contested issue in most receiving countries. Indeed, the reaction to immigrants has not been the same in all countries: There are substantial differences in the timing and extent of the politicisation of the issue of immigration. Even where we observe similar migration patterns, policy responses and societal tensions associated with migrant groups seem to differ noticeably.
The SOM project seeks explanations to the cross-country differences in the politicisation of immigration and the integration of immigrants. It compares seven Western European countries that differ institutionally, as well as in the number and nature of immigrants they received. The process of politicisation is described in terms of salience of the issues of immigration and integration, as well as in terms of polarisation of opinions around those issues. High salience implies that immigration is talked about, whereas polarisation refers to differences in opinions. Immigration is considered politicised when high levels of salience and polarisation are observed at the same time.
The aims of the project include:
1. increasing knowledge about conflicts over the social and political participation of immigrants in Western Europe;
2. determining why and when potential conflicts become politicised examining both anti-immigration and anti-racist movements; and
3. increasing knowledge of how institutional conditions constrain processes of politicisation.
The output of the project includes a large longitudinal cross-country dataset of political claims. Newspaper articles from 1995 to 2009 were chosen to observe political claims claims-making. A political claim describes any public political activity such as a press statement, a protest or a policy initiative by which political actors call for change in policy. This dataset allows exploring how and when immigration and integration become politicised. In addition the SOM project gathered data as explanatory factors: extensive demographic data, information on the legal situation and immigration policies, as well as institutional factors related to the electoral and party systems of the countries.
In comparative analyses the project found that across all seven countries, the salience of immigration was relatively low in 1995. It increased in the early 2000s to decrease again somewhat in the late 2000s. Within this pattern, there are large variations from year to year, but these increases and decreases in salience differ from country to country. Immigration and integration appears least salient in Ireland and Wallonia, and most in the Netherlands and Austria. At the same time, immigration is a more contested issue in certain countries than others. The United Kingdom (UK) displays the highest level of polarisation over time, the German speaking part of Switzerland the lowest. It appears that in countries where party politics are dominated by two large parties (UK and Spain), the issue of immigration is most polarised.
Neither the salience nor the polarisation of the issue are directly related to the influx of immigrants, the share of foreign born residents, the policy responses, or the state of the economy. Indeed, country specific factors affect polarisation and salience of immigration. In all countries politicisation is a mixture of political leadership or initiative, and circumstances that provide opportunities to influence politics. In most countries politicisation is driven by political parties. This is less pronounced in Ireland, the UK, and Switzerland, where the process of politicisation often is initiated by civil society actors and journalists. Systemic features of these countries may explain these differences.
Project context and objectives:
The aim of the comparative project SOM is to determine why and when potential conflicts over migration become politicised. The guiding research question is: 'What causes large cross-national differences in the ways the presence of immigrants and their full political and social participation becomes contested and politicised'?
The project pursued four avenues in order to address the research question. First, the project partners seek to contribute to the contemporary literature on political debates and policies about migration and integration. In particular they project increases knowledge about the conflicts over the social and political participation of immigrants in Western Europe.
Second, despite the fact that conflicts are at the heart of political sciences, little is known about?issue formation. 'Why and when do potential conflicts become politicised, and when and why do they not become politicised'? By addressing these questions, the project is relevant to a much broader group of researchers than those interested in immigration and interaction policies.
Third, the project increases knowledge of the way political processes are constrained by institutional conditions. By including a temporal dimension in the research design, the project makes strides toward better understanding the involved dynamics. Fourth, the project provides policy-relevant information by assessing which actions of state intuitions have been more or less successful in managing conflict on immigration and integration. From the point of view of policy-making, the fact that immigration becomes a contested issue is not necessarily a ?political? or ?social? problem, but may be just part of ordinary policy processes in which opposite views emerge around how to manage society. However, when contestation transforms into conflict, and especially violent conflict, social cohesion might be at risk. It is therefore, in our opinion, relevant for policy-makers to learn which institutions might promote heightened conflict and which might reduce it to reasonable contestation.
To better understand the dynamics of politicisation, the project covers countries with different political reactions to immigration. It includes two countries where immigration started relatively late (Spain and Ireland), three countries where new or established parties successfully mobilised support on the anti-immigration issue (Switzerland, Belgium notably in Flanders, and Austria), one country with many immigrants, yet without giving rise to successful anti-immigration parties (UK), and one country where such parties have been around for a long time, but only very recently with some electoral success (the Netherlands). The inter-country variation allows to explain overarching trends and to identify which systemic elements contribute to determining politicisation.
The SOM project adopted a variety of methodological approaches mixing quantitative and qualitative techniques. A quantitative content analysis of newspapers over a period of 15 years (1995-2009) plays a central role in the project. Specifically, political claims? made in the media were sampled, coded and measured over time and across countries. To identify politicisation, the study made use of a random sample of newspapers, consisting of 700 days per country across two different newspapers (one tabloid and one broadsheet). In Belgium and Switzerland we accounted for the presence of linguistic regions by including two newspapers of the major language regions respectively. The project recognises two aspects of politicisation, namely salience and polarisation. Salience was measured in terms of number of claims on the issue per sampled day. This provides an average number of claims that is expressed here in yearly terms, by employing of moving averages. Moving averages are particularly suitable to describe long term trends. Polarisation is measured, employing a 5-point scale, on whether the relevant political claim is displaying a position favourable or hostile towards migrants.
Project results:
The SOM project seeks to explain cross-country differences in the politicisation of the migration issue. It compares seven Western European countries that differ institutionally, as well as in the number of immigrants they received. The process of politicisation is described in terms of salience of the issues of immigration and integration, as well as in terms of polarisation of opinions around those issues. Immigration is considered properly politicised when high levels of salience and polarisation are observed at the same time: many claims are made about immigration and integration, and political actors take different views on the issue.
Substantial numbers of immigrants arrived in Western Europe since the 1960s. At least from the mid-Nineties onwards, the presence and integration of immigrants has become a contested issue in all but a few receiving countries. However, there are substantial differences between countries in the timing and extent of the politicisation of the issue of immigration, despite similar migration patterns, policy responses and possible societal tensions associated with migrant groups.
The output of the project includes a large longitudinal (1995-2009) cross-country dataset of political claims. Political claims are public political activities such as press statements, protests or policy initiatives by various political actors, and are observed via newspapers. This dataset allows exploring how and when immigration and integration become politicised. In addition, the SOM-project gathered demographic and legal data on immigration and civic integration to be used as explanatory factors. Furthermore, information on the political system such as their electoral and party systems of the countries under study is integrated in the analysis.
The key findings of the project are as follows:
First, across all seven countries, the salience of immigration and integration was relatively low in 1995. It increased in the early 2000s and in most countries decreased somewhat in the late 2000s. An exception Is Austria, where the salience of immigration and integration increased toward the end of the period covered. The salience of immigration and integration is expressed as the relative number of political claims found on a fixed number of newspaper editions around randomly sampled days.
Second, there are country-specific trends. Across countries, the issue is least salient in Ireland and Wallonia, and most in the Netherlands, Austria and Spain. All countries experience peaks of politicisation at different points of time. In the Netherlands peaks follow traumatic events as the murders of Pim Fortuyn and Theo van Gogh, but after the shocks, the saliency of the issue returns rapidly to much lower levels. The peaks in Spain are not related to such events; while Austria is the only country where there has been a clear upward trend in salience throughout the period covered.
Third, the debate has shifted from questions of immigration to questions of integration. Within the overall theme of the project, we distinguished between claims made pertaining to migration and claims pertaining to civic integration. The proportion of claims about migration decreased relative to the proportion of claims about migrant integration. This trend is fairly uniform across countries. Ireland and Austria stand out as countries where asylum seekers are more politicised than in the other countries under study.
Fourth, immigration is a more contested issue in certain countries than others. The United Kingdom displays the highest level of polarisation over time, the German speaking part of Switzerland the lowest. There are substantial fluctuations over time, but countries where party politics is dominated by two large parties (UK and Spain) tend to be the most polarised. The German speaking part of Switzerland, which displays an over-time increase in terms of salience, surprisingly presents low polarisation scores. As such, we hardly find a simultaneous presence of high salience and high polarisation, which constitutes the basis for politicisation. Similarly, the two Belgian regions show very similar degrees of polarisation. Combining salience and polarisation, Flanders displays higher degrees of politicisation of immigration than Wallonia.
In the Netherlands, by contrast, the issue of immigration appears to have been latently salient but not strongly polarised over time. The peaks in salience in 2001 / 2002 and 2004 are not reflected in high polarisation. In Ireland, immigration displays the lowest degree of saliency, and it appears to have been an issue on which no strongly polarised positions were voiced. In the United Kingdom, a high level of polarisation is registered, in fact the highest in the group of countries under study here. Polarisation is relatively stable in the late 1999s and early 2000, and it then fluctuates quite a bit. However, values remain on the high side of the spectrum, showing that political actors tend to voice quite opposed positions on the issue of immigration. In Austria, polarisation does not fluctuate very much. Peaks are present in 1996 and in 2004, but the overall tendency seems to be characterised by little polarisation. Spain, despite the lack of anti-immigrant parties at the national level, does show some degrees of polarisation, with the second highest average of the countries studied in the project. All values fluctuate very little over time.
Surprisingly, the radical right parties are not the main claimant at the level of parties. This finding is particularly strong in Austria, and with reference to the Freedom Party (FPÖ). It is the mainstream parties that most frequently appear as claimants on the immigration issue. This difference does not change during the FPÖ?s participation in federal government (2000-2005), or for the Swiss People's Party (SVP) which is part of government. At the other end of the spectrum, immigrant actors and organisations play virtually no role as claimants in the politicisation of immigration. Claims-making on immigration in Austria is first and foremost a discussion about and of immigrants, not by immigrants themselves.
Fifth, salience and polarisation of immigration and integration are not clearly related to the influx of immigrants, to the share of foreign-born residents, the policy responses, or the state of the economy. There is no systematic relationship between the number of immigrants in a country or the number of immigrants entering a country and the extent to which the issue is politicised. This remains true when we distinguish between different kinds of immigrants coming from different parts of the world. Moreover, there is also no direct relationship between our indicators of politicisation and measures of the state of the economy, such as growth in Gross domestic product (GDP). The issue of immigration does not ?automatically? become politicised once there are a large number of immigrants.
Sixth, country-specific factors affect polarisation and salience of immigration. In all countries, politicisation is a mixture of political leadership or initiative, and circumstances that provide opportunities to influence politics. There are, however, large differences between the countries in the extent to which ministers and parliamentarians dominate the debate on immigration and civic integration. In most countries politicisation is driven by political parties (top-down), but in Ireland, the UK and to a lesser extent Switzerland, there is more room for civil society actors and journalists (bottom-up). In Ireland and the UK, parties are much less prominent in the debate than in the other countries. Systemic features of these countries may explain these differences.
Preliminary analysis shows that in all countries the mainstream parties, who in the past or currently hold government responsibility, have an incentive not to politicise the issue. Yet, they were largely unsuccessful in doing so. In countries with multiparty systems (Belgium, Austria, Switzerland and the Netherlands), it was relatively easy for radical right parties to form and to challenge the mainstream parties on this sensitive issue. Such politicisation often took place in these countries, referendums were often the catalysing events. In Belgium party actors were slightly less dominant in the debate than in the other multi-party systems. A tentative explanation is that the cordon sanitaire against the anti-immigration parties Vlaams Belang and Front National enabled the mainstream parties to avoid the debate on immigration related issues.
In the UK, where party politics is dominated by two large parties, politicisation on the issue did not take place in the party arena as much as in the other countries. Due to the electoral system, the mainstream party has more opportunities to keep the issue off the party agenda. Spain is also mostly dominated by two parties. Here, however, governments could not keep the issue off the agenda, when large numbers of boat people from Western Africa arrived on the Canary Islands. Yet, perhaps interestingly, the parties de-politicised the issue around the Madrid bombing in 2004. In Ireland the issue was hardly politicised throughout the period. This had partially to do with the specific structure of the Irish party system, where the government party took the most right-wing positions on immigration of all parties, thus leaving no room for the opposition or for an anti-immigration party to form. In addition to knowledge, the project produced large datasets. These data are freely available from the project dataverse: http://dvn.iq.harvard.edu/dvn/dv/som. This Harvard-backed data repository is widely recognised in academic circles, and allows convenient data sharing in multiple file formats. Basic descriptive analyses can be carried out online. The dataset from the media analysis will be added to the Dataverse within a year of the project end. All documentation, including the codebook and reliability analyses, has already been made available to other researchers by publishing them as SOM working papers. All SOM Working Papers are distributed through the Social Science Research Network (http://ssrn.com) to reach a wider audience. In addition to the working papers, the project partners were actively engaged on several conferences. The first journal articles are under review and around eight are near completion. A book is in preparation.
Potential impact:
The rise of anti-immigrant movements and politicisation of immigration mean that the research produced by the project is topical. A state-of-the-art website was set up at the beginning of the project and regularly kept updated. Full description of project goals, structure, events, and consortium participants are listed. All working papers, policy briefs, and published data are available from the website. Users can subscribe to the project newsletter. The website will be maintained beyond the formal end of the project to inform about publications and other relevant activities.
The project partners have made available all deliverables that are of general or specialist interest as working papers. To reach a wider audience, the working papers were disseminated using the Social Science Research Network (http://ssrn.com). Some of the working papers were in the top-10 of relevant categories, indicating interest among academics. Academic research beyond these deliverables was undertaken, and several publications are close to submission to top journals. Most of these papers are jointly authored, many of which by researchers who did not know each other before the beginning of the project. An edited book on the countries in the study is in preparation.
Three policy briefs were written to present the empirical findings of the project and reflect on the policy implications. These are available on the project website, and were distributed to interested individuals through the SOM newsletter and personal networks. Members of the advisory board work in policy-related areas and were keen to disseminate the findings of the project. At the final conference in Brussels and at national dissemination events the project partners took care to disseminate research findings beyond the academic setting to policymakers, practitioners, and interested stakeholders.
The research was internally reviewed at SOM workshops, but also externally at academic conferences. This happened both within the IMISCOE network as well as in international conferences. Future collaboration in the framework of Horizon 2020 is strongly considered, and the project partners have already developed joint proposals on related issues. A privately funded project at the Université libre de Bruxelles is under way to study and analyse the recent shift towards more restrictive integration policies. The idea is to reproduce the analytical design of SOM, based mostly upon an in-depth analysis of claims in the media, in order to see whether we can trace back what are the sources of the recent shift observed in several EU countries towards more restrictive integration policies. A joint project by the Universities of Neuchâtel and Vienna was submitted to SNSF and FSW to study the politicisation of and mobilisation against asylum seekers. The partners in Brussels and Dublin collaborate on a joint research project on the personal vote in PR list systems. They are collaborating on the use of the mock-ballot technique for the elections studies conducted in Belgium in 2012 and 2014. The project is financed by the FNRS and the Belgian Science Foundation (BelSpo). Without SOM, these collaborations would certainly not have existed.
List of websites: http://www.som-project.eu
Since the 1960s, substantial numbers of immigrants migrated to Western Europe. At least from the mid-Nineties onwards, the presence and integration of immigrants became a contested issue in most receiving countries. Indeed, the reaction to immigrants has not been the same in all countries: There are substantial differences in the timing and extent of the politicisation of the issue of immigration. Even where we observe similar migration patterns, policy responses and societal tensions associated with migrant groups seem to differ noticeably.
The SOM project seeks explanations to the cross-country differences in the politicisation of immigration and the integration of immigrants. It compares seven Western European countries that differ institutionally, as well as in the number and nature of immigrants they received. The process of politicisation is described in terms of salience of the issues of immigration and integration, as well as in terms of polarisation of opinions around those issues. High salience implies that immigration is talked about, whereas polarisation refers to differences in opinions. Immigration is considered politicised when high levels of salience and polarisation are observed at the same time.
The aims of the project include:
1. increasing knowledge about conflicts over the social and political participation of immigrants in Western Europe;
2. determining why and when potential conflicts become politicised examining both anti-immigration and anti-racist movements; and
3. increasing knowledge of how institutional conditions constrain processes of politicisation.
The output of the project includes a large longitudinal cross-country dataset of political claims. Newspaper articles from 1995 to 2009 were chosen to observe political claims claims-making. A political claim describes any public political activity such as a press statement, a protest or a policy initiative by which political actors call for change in policy. This dataset allows exploring how and when immigration and integration become politicised. In addition the SOM project gathered data as explanatory factors: extensive demographic data, information on the legal situation and immigration policies, as well as institutional factors related to the electoral and party systems of the countries.
In comparative analyses the project found that across all seven countries, the salience of immigration was relatively low in 1995. It increased in the early 2000s to decrease again somewhat in the late 2000s. Within this pattern, there are large variations from year to year, but these increases and decreases in salience differ from country to country. Immigration and integration appears least salient in Ireland and Wallonia, and most in the Netherlands and Austria. At the same time, immigration is a more contested issue in certain countries than others. The United Kingdom (UK) displays the highest level of polarisation over time, the German speaking part of Switzerland the lowest. It appears that in countries where party politics are dominated by two large parties (UK and Spain), the issue of immigration is most polarised.
Neither the salience nor the polarisation of the issue are directly related to the influx of immigrants, the share of foreign born residents, the policy responses, or the state of the economy. Indeed, country specific factors affect polarisation and salience of immigration. In all countries politicisation is a mixture of political leadership or initiative, and circumstances that provide opportunities to influence politics. In most countries politicisation is driven by political parties. This is less pronounced in Ireland, the UK, and Switzerland, where the process of politicisation often is initiated by civil society actors and journalists. Systemic features of these countries may explain these differences.
Project context and objectives:
The aim of the comparative project SOM is to determine why and when potential conflicts over migration become politicised. The guiding research question is: 'What causes large cross-national differences in the ways the presence of immigrants and their full political and social participation becomes contested and politicised'?
The project pursued four avenues in order to address the research question. First, the project partners seek to contribute to the contemporary literature on political debates and policies about migration and integration. In particular they project increases knowledge about the conflicts over the social and political participation of immigrants in Western Europe.
Second, despite the fact that conflicts are at the heart of political sciences, little is known about?issue formation. 'Why and when do potential conflicts become politicised, and when and why do they not become politicised'? By addressing these questions, the project is relevant to a much broader group of researchers than those interested in immigration and interaction policies.
Third, the project increases knowledge of the way political processes are constrained by institutional conditions. By including a temporal dimension in the research design, the project makes strides toward better understanding the involved dynamics. Fourth, the project provides policy-relevant information by assessing which actions of state intuitions have been more or less successful in managing conflict on immigration and integration. From the point of view of policy-making, the fact that immigration becomes a contested issue is not necessarily a ?political? or ?social? problem, but may be just part of ordinary policy processes in which opposite views emerge around how to manage society. However, when contestation transforms into conflict, and especially violent conflict, social cohesion might be at risk. It is therefore, in our opinion, relevant for policy-makers to learn which institutions might promote heightened conflict and which might reduce it to reasonable contestation.
To better understand the dynamics of politicisation, the project covers countries with different political reactions to immigration. It includes two countries where immigration started relatively late (Spain and Ireland), three countries where new or established parties successfully mobilised support on the anti-immigration issue (Switzerland, Belgium notably in Flanders, and Austria), one country with many immigrants, yet without giving rise to successful anti-immigration parties (UK), and one country where such parties have been around for a long time, but only very recently with some electoral success (the Netherlands). The inter-country variation allows to explain overarching trends and to identify which systemic elements contribute to determining politicisation.
The SOM project adopted a variety of methodological approaches mixing quantitative and qualitative techniques. A quantitative content analysis of newspapers over a period of 15 years (1995-2009) plays a central role in the project. Specifically, political claims? made in the media were sampled, coded and measured over time and across countries. To identify politicisation, the study made use of a random sample of newspapers, consisting of 700 days per country across two different newspapers (one tabloid and one broadsheet). In Belgium and Switzerland we accounted for the presence of linguistic regions by including two newspapers of the major language regions respectively. The project recognises two aspects of politicisation, namely salience and polarisation. Salience was measured in terms of number of claims on the issue per sampled day. This provides an average number of claims that is expressed here in yearly terms, by employing of moving averages. Moving averages are particularly suitable to describe long term trends. Polarisation is measured, employing a 5-point scale, on whether the relevant political claim is displaying a position favourable or hostile towards migrants.
Project results:
The SOM project seeks to explain cross-country differences in the politicisation of the migration issue. It compares seven Western European countries that differ institutionally, as well as in the number of immigrants they received. The process of politicisation is described in terms of salience of the issues of immigration and integration, as well as in terms of polarisation of opinions around those issues. Immigration is considered properly politicised when high levels of salience and polarisation are observed at the same time: many claims are made about immigration and integration, and political actors take different views on the issue.
Substantial numbers of immigrants arrived in Western Europe since the 1960s. At least from the mid-Nineties onwards, the presence and integration of immigrants has become a contested issue in all but a few receiving countries. However, there are substantial differences between countries in the timing and extent of the politicisation of the issue of immigration, despite similar migration patterns, policy responses and possible societal tensions associated with migrant groups.
The output of the project includes a large longitudinal (1995-2009) cross-country dataset of political claims. Political claims are public political activities such as press statements, protests or policy initiatives by various political actors, and are observed via newspapers. This dataset allows exploring how and when immigration and integration become politicised. In addition, the SOM-project gathered demographic and legal data on immigration and civic integration to be used as explanatory factors. Furthermore, information on the political system such as their electoral and party systems of the countries under study is integrated in the analysis.
The key findings of the project are as follows:
First, across all seven countries, the salience of immigration and integration was relatively low in 1995. It increased in the early 2000s and in most countries decreased somewhat in the late 2000s. An exception Is Austria, where the salience of immigration and integration increased toward the end of the period covered. The salience of immigration and integration is expressed as the relative number of political claims found on a fixed number of newspaper editions around randomly sampled days.
Second, there are country-specific trends. Across countries, the issue is least salient in Ireland and Wallonia, and most in the Netherlands, Austria and Spain. All countries experience peaks of politicisation at different points of time. In the Netherlands peaks follow traumatic events as the murders of Pim Fortuyn and Theo van Gogh, but after the shocks, the saliency of the issue returns rapidly to much lower levels. The peaks in Spain are not related to such events; while Austria is the only country where there has been a clear upward trend in salience throughout the period covered.
Third, the debate has shifted from questions of immigration to questions of integration. Within the overall theme of the project, we distinguished between claims made pertaining to migration and claims pertaining to civic integration. The proportion of claims about migration decreased relative to the proportion of claims about migrant integration. This trend is fairly uniform across countries. Ireland and Austria stand out as countries where asylum seekers are more politicised than in the other countries under study.
Fourth, immigration is a more contested issue in certain countries than others. The United Kingdom displays the highest level of polarisation over time, the German speaking part of Switzerland the lowest. There are substantial fluctuations over time, but countries where party politics is dominated by two large parties (UK and Spain) tend to be the most polarised. The German speaking part of Switzerland, which displays an over-time increase in terms of salience, surprisingly presents low polarisation scores. As such, we hardly find a simultaneous presence of high salience and high polarisation, which constitutes the basis for politicisation. Similarly, the two Belgian regions show very similar degrees of polarisation. Combining salience and polarisation, Flanders displays higher degrees of politicisation of immigration than Wallonia.
In the Netherlands, by contrast, the issue of immigration appears to have been latently salient but not strongly polarised over time. The peaks in salience in 2001 / 2002 and 2004 are not reflected in high polarisation. In Ireland, immigration displays the lowest degree of saliency, and it appears to have been an issue on which no strongly polarised positions were voiced. In the United Kingdom, a high level of polarisation is registered, in fact the highest in the group of countries under study here. Polarisation is relatively stable in the late 1999s and early 2000, and it then fluctuates quite a bit. However, values remain on the high side of the spectrum, showing that political actors tend to voice quite opposed positions on the issue of immigration. In Austria, polarisation does not fluctuate very much. Peaks are present in 1996 and in 2004, but the overall tendency seems to be characterised by little polarisation. Spain, despite the lack of anti-immigrant parties at the national level, does show some degrees of polarisation, with the second highest average of the countries studied in the project. All values fluctuate very little over time.
Surprisingly, the radical right parties are not the main claimant at the level of parties. This finding is particularly strong in Austria, and with reference to the Freedom Party (FPÖ). It is the mainstream parties that most frequently appear as claimants on the immigration issue. This difference does not change during the FPÖ?s participation in federal government (2000-2005), or for the Swiss People's Party (SVP) which is part of government. At the other end of the spectrum, immigrant actors and organisations play virtually no role as claimants in the politicisation of immigration. Claims-making on immigration in Austria is first and foremost a discussion about and of immigrants, not by immigrants themselves.
Fifth, salience and polarisation of immigration and integration are not clearly related to the influx of immigrants, to the share of foreign-born residents, the policy responses, or the state of the economy. There is no systematic relationship between the number of immigrants in a country or the number of immigrants entering a country and the extent to which the issue is politicised. This remains true when we distinguish between different kinds of immigrants coming from different parts of the world. Moreover, there is also no direct relationship between our indicators of politicisation and measures of the state of the economy, such as growth in Gross domestic product (GDP). The issue of immigration does not ?automatically? become politicised once there are a large number of immigrants.
Sixth, country-specific factors affect polarisation and salience of immigration. In all countries, politicisation is a mixture of political leadership or initiative, and circumstances that provide opportunities to influence politics. There are, however, large differences between the countries in the extent to which ministers and parliamentarians dominate the debate on immigration and civic integration. In most countries politicisation is driven by political parties (top-down), but in Ireland, the UK and to a lesser extent Switzerland, there is more room for civil society actors and journalists (bottom-up). In Ireland and the UK, parties are much less prominent in the debate than in the other countries. Systemic features of these countries may explain these differences.
Preliminary analysis shows that in all countries the mainstream parties, who in the past or currently hold government responsibility, have an incentive not to politicise the issue. Yet, they were largely unsuccessful in doing so. In countries with multiparty systems (Belgium, Austria, Switzerland and the Netherlands), it was relatively easy for radical right parties to form and to challenge the mainstream parties on this sensitive issue. Such politicisation often took place in these countries, referendums were often the catalysing events. In Belgium party actors were slightly less dominant in the debate than in the other multi-party systems. A tentative explanation is that the cordon sanitaire against the anti-immigration parties Vlaams Belang and Front National enabled the mainstream parties to avoid the debate on immigration related issues.
In the UK, where party politics is dominated by two large parties, politicisation on the issue did not take place in the party arena as much as in the other countries. Due to the electoral system, the mainstream party has more opportunities to keep the issue off the party agenda. Spain is also mostly dominated by two parties. Here, however, governments could not keep the issue off the agenda, when large numbers of boat people from Western Africa arrived on the Canary Islands. Yet, perhaps interestingly, the parties de-politicised the issue around the Madrid bombing in 2004. In Ireland the issue was hardly politicised throughout the period. This had partially to do with the specific structure of the Irish party system, where the government party took the most right-wing positions on immigration of all parties, thus leaving no room for the opposition or for an anti-immigration party to form. In addition to knowledge, the project produced large datasets. These data are freely available from the project dataverse: http://dvn.iq.harvard.edu/dvn/dv/som. This Harvard-backed data repository is widely recognised in academic circles, and allows convenient data sharing in multiple file formats. Basic descriptive analyses can be carried out online. The dataset from the media analysis will be added to the Dataverse within a year of the project end. All documentation, including the codebook and reliability analyses, has already been made available to other researchers by publishing them as SOM working papers. All SOM Working Papers are distributed through the Social Science Research Network (http://ssrn.com) to reach a wider audience. In addition to the working papers, the project partners were actively engaged on several conferences. The first journal articles are under review and around eight are near completion. A book is in preparation.
Potential impact:
The rise of anti-immigrant movements and politicisation of immigration mean that the research produced by the project is topical. A state-of-the-art website was set up at the beginning of the project and regularly kept updated. Full description of project goals, structure, events, and consortium participants are listed. All working papers, policy briefs, and published data are available from the website. Users can subscribe to the project newsletter. The website will be maintained beyond the formal end of the project to inform about publications and other relevant activities.
The project partners have made available all deliverables that are of general or specialist interest as working papers. To reach a wider audience, the working papers were disseminated using the Social Science Research Network (http://ssrn.com). Some of the working papers were in the top-10 of relevant categories, indicating interest among academics. Academic research beyond these deliverables was undertaken, and several publications are close to submission to top journals. Most of these papers are jointly authored, many of which by researchers who did not know each other before the beginning of the project. An edited book on the countries in the study is in preparation.
Three policy briefs were written to present the empirical findings of the project and reflect on the policy implications. These are available on the project website, and were distributed to interested individuals through the SOM newsletter and personal networks. Members of the advisory board work in policy-related areas and were keen to disseminate the findings of the project. At the final conference in Brussels and at national dissemination events the project partners took care to disseminate research findings beyond the academic setting to policymakers, practitioners, and interested stakeholders.
The research was internally reviewed at SOM workshops, but also externally at academic conferences. This happened both within the IMISCOE network as well as in international conferences. Future collaboration in the framework of Horizon 2020 is strongly considered, and the project partners have already developed joint proposals on related issues. A privately funded project at the Université libre de Bruxelles is under way to study and analyse the recent shift towards more restrictive integration policies. The idea is to reproduce the analytical design of SOM, based mostly upon an in-depth analysis of claims in the media, in order to see whether we can trace back what are the sources of the recent shift observed in several EU countries towards more restrictive integration policies. A joint project by the Universities of Neuchâtel and Vienna was submitted to SNSF and FSW to study the politicisation of and mobilisation against asylum seekers. The partners in Brussels and Dublin collaborate on a joint research project on the personal vote in PR list systems. They are collaborating on the use of the mock-ballot technique for the elections studies conducted in Belgium in 2012 and 2014. The project is financed by the FNRS and the Belgian Science Foundation (BelSpo). Without SOM, these collaborations would certainly not have existed.
List of websites: http://www.som-project.eu