CROBOCOB CROSS-BORDER COMMUTING IN THE EU: OBSTACLES AND BARRIERS FINAL REPORT

Transcription

1 CROBOCOB CROSS-BORDER COMMUTING IN THE EU: OBSTACLES AND BARRIERS FINAL REPORT 1

2 SCIENTIFIC REPORT Abstract This accompanying measure reviewed existing and collected new evidence of crossborder commuting in six border regions in the EU. The study investigated reasons for immobility and the perception of factors which inhibit potential commuting to work across borders. It produced six detailed border reports discussing labour market characteristics and policies, evidence of the level and type of cross-border commuting and of factors inhibiting cross-border commuting. Subsequently, 102 interviews were conducted on four border regions (Germany/Netherlands, Sweden/Denmark, Finland/Sweden and Italy/France). The results in brief: High levels of indifference (interviewees not even considering the option of cross-border commuting) were found in France, the Netherlands and also Denmark, in contrast to the Italian and German samples, as well as either side of the Finnish- Swedish border. Some borders hence demonstrated a remarkable cross-border imbalance, as far as levels of indifference are concerned. The relatively better domestic labour market seems to be a relevant (stay) factor. However, less favourable employment prospects are clearly not sufficient for crossborder commuting. The scale of existing commuting, personal contacts with crossborder commuters and readily available information are important ingredients for a critical mass which seems to be required to counteract preconceptions which prevent job seekers to contemplate cross-border commuting. The interest in working abroad (or perceived obstacles involved) often differs within national samples. For example, within the German sample, potential commuters could be distinguished by different qualification and skill backgrounds who made use of different channels of information about work opportunities in the Netherlands. Lack of information about jobs and also about possible consequences of working opportunities at the other side of the border was a very significant barrier. Job adverts in local newspapers but also a better knowledge of local EURES offices or similar work agencies could improve the potential for cross-border commuting. Cultural aspects are relevant, but often appeared as less important than other factors (lack of job information; transport costs and time). The inability to speak the language of the other side of the border sufficiently is clearly perceived to be a problem, but not for all groups or for all types of jobs. Complications with tax and social security regulations were mentioned by those with previous experience with cross-border commuting, in contrast with many potential commuters, who play down any possible problems of this sort. The results illustrate relevance of the research for policies intent on raising the level of labour mobility across EU member states. However, more empirical 2

3 evidence is needed which would support the notion of a critical mass of factors which either reduce or enhance the propensity for cross-border commuting. 3

4 1. Executive summary Free mobility of labour is one of the cornerstones of the European Union. And yet, compared with the USA, there is relatively little labour mobility across member states. What are the reasons for this? Existing research on paid work performed by EU citizens in member states other than their own has concentrated on labour migration, and tended to be quantitative in nature, approaching the topic from an economic perspective seeking explanations of cross border commuting in terms of utility maximisation. By comparison, there is very little in-depth qualitative and comparative knowledge about what factors might be perceived by potential commuters as barriers and obstacles to seeking paid work on the other side of the border. Border influences can be divided into objective physical and institutional constraints (hard factors), and subjective values and meanings attached to those as well as to other potential factors, such as linguistic, cultural or differences (soft factors). This smallscale Accompanying Measure sought to tap into both of these by exploring the border as 'a mental map'. The study collected and reviewed existing evidence in the field, followed by 102 interviews in four border regions. On the basis of a geographical spread and the diverse volume of flows of cross-border commuting to work, six borders were investigated: Netherlands/Belgium, Germany/Netherlands, Denmark/Germany; Sweden/Denmark, Finland/Sweden and Italy/France. The first stage of the study produced six border reports featuring labour market characteristics, labour market policies, evidence of the level and type of cross-border commuting and of factors inhibiting cross-border commuting. Common characteristics were identified, such as the relative disadvantage of young people without skills or qualifications as well as older workers with either obsolete or few skills. All of the countries have experienced an increasing emphasis upon activation policies for the unemployed combined with attempts at welfare state reform and change. But there are also considerable differences across the selected border regions, concerning geography, population density, transport links, institutional cross-border cooperation and language. The borders include sparsely populated areas (Sweden/Finland) and relatively densely populated borders (e.g. Netherlands/Germany/Belgium). The German/Dutch border has a long history of cross-border cooperation, and is very well researched, compared to the more recent institutional forms which developed on the Swedish/Finnish or French/Italian border. There are examples of high unemployment on both sides of the border (Finland/Sweden), relatively low unemployment on both sides (Sweden/Denmark) and also contrasting levels of unemployment (the Netherlands/Germany, Italy/France, Germany/Denmark). The latter, combined with the differences in property prices, housing tenure and systems of housing taxation, might be seen as being a major factor in cross-border flows if cross-border commuting is seen solely as a form of economically rational behaviour. Linguistic differences are also contrasting in these border regions. On the one hand there is little difference on the Belgium/Netherlands border contrasting with the existence of more than two languages in regular use on the Swedish/Finnish border. Language may be perceived as a barrier on one side of the border for commuters but may not be perceived as such a problem the other way around. 4

5 The study proceeded with interviews of potential commuters in four regions of national borders (Finland/Sweden; Denmark/Sweden, Germany/Netherlands and Italy/France). The scale of the study did not allow to interview representative samples. However, common guidelines applied with regard to age limits, gender balance, family situation and type of job seekers (registered and not registered), taking account of the characteristics of existing commuters in each region. Indifference was identified by the research teams as a key concept to guide the interviews and analysis. The eight subsamples allowed a rough categorisation between respondents who had at least contemplated commuting to work across the border (and perhaps rejected it) and those who typically stated that the idea had not occurred to them (the indifferent). The results in brief: High levels of indifference (not considering the option of cross-border commuting) were found in France, the Netherlands and also Denmark, in contrast to the Italian and German samples, as well as either side of the Finnish- Swedish border. Some borders hence demonstrated a remarkable cross-border imbalance, as far as the threshold of indifference is concerned. This is most obvious at the Italian-French border where almost none of the French interviewees even considered the idea of seeking work in Italy. Where there is a high level of indifference, the relatively better domestic labour market seems to be a dominant (stay) factor which inhibits a more active engagement with the idea of working at the other side of the border. By contrast, a low level of indifference seems to be prevalent in regions with a domestic lack of labour demand combined with a perception of better job (or pay) prospects across the border. However, a less favourable employment situation is clearly not sufficient for cross-border commuting, and often not even for considering the idea. The level of existing cross-border commuting, personal knowledge of workers already commuting and readily available information about opportunities (e.g. job adverts in local newspapers but also the existence and knowledge of EURES or other special work agencies) are three important ingredients of a critical mass which are needed to counteract preconceptions which prevent job seekers to contemplate cross-border commuting. Within national samples, sub-groups could often be distinguished. They suggest that the interest in working abroad (or perceived obstacles involved) differ in accordance with educational, occupational and sometimes socio-cultural backgrounds. This is most obvious within the Italian sample. The French labour market is regarded as attractive for less qualified manual workers, but as closed for better qualified, skilled service sector employees. With the German sample, potential commuters could be distinguished by different qualification and skill backgrounds who made use of different channels of information about working in the Netherlands. Factors commonly perceived as barriers by respondents as a whole can be grouped around a number of themes: knowledge; cultural/social; personal/family; institutional and infrastructural. 5

6 Lack of information about jobs and also about possible consequences of working opportunities at the other side of the border was a very significant barrier mentioned. Here, job adverts from across the border in local newspapers but also a better knowledge of the services of local EURES offices or similar work agencies could improve the potential for cross-border commuting. Cultural factors (organisational cultures and working environment; social reference points; perceived attitudes towards foreign workers) are relevant, and in some border regions (Italy-France) apparently a very important barrier to commuting. However, having to confront an unfamiliar cultural context appeared often as less important than other factors (lack of job information in particular; transport problems). Even where cultural stay factors were mentioned, such as in Denmark, other factors (such as lack of information or the cost of commuting) seemed equally or even more important. The inability to speak the language of the other side of the border sufficiently is clearly perceived to be a problem, but not for all groups or for all types of jobs. For some Germans, for example, not being able to speak Dutch did not figure as an obstacle because of the type of jobs sought (language skills not needed) or the perceptions that most Dutch speak German. The significance of physical factors which inhibit cross-border commuting varied considerably. Problems of cost and time of travel, for example, were important in some Finnish and Swedish locations, and were mentioned also at both sides of the Øresund, where crossing borders seems less of a problem than the cost and time of commuting per se. Interestingly, hard factors, such as potential complications with tax and social security regulations, and thus possible problems establishing rights to income maintenance played a rather minor role. However, many respondents with previous experience of cross-border commuting complained about difficulties of claiming sick pay, maternity benefits or pensions, which were partly attributed to insufficient information or expertise on the part of employment offices, social insurance agencies or trade unions. There seems to be a contrast with many potential commuters, who play down any possible complications of this sort. The above illustrates the policy relevance of this exploratory project. Beyond a listing of potential barriers to cross-border commuting, the results indicate how factors might interact with each other and which barriers are perceived as relevant in which border region (or side of the same border). More conceptual work regarding the key concept of indifference is required, as is more empirical evidence which would support the notion of a critical mass of factors which either limit or enhance the propensity for crossborder commuting. But even this small-scale study can inform policies aimed at increasing labour mobility. The data provided indicates that local EURES offices can play an important role in enhancing the propensity of cross-border employment. However, their effectiveness is context-specific and dependent on other factors which need to be taken account of in policy formulation. A full copy of the results will be submitted to EURES and other relevant organisations at European and Member State level. The preliminary findings and hypotheses generated by this Accompanying Measure will inform further comparative research and will involve more network members, including colleagues from accession countries. 6

7 2. Background and objectives of the project There are about 5.5 million nationals of the EU living or working in a member state other than their own, and 15.5 million nationals of non-eu states (Vandamme, 2000). Most of the former group are people who have migrated from one to another country, but a minority cross the border as commuters. While a clear-cut distinction between cross-border commuting and labour migration is difficult to make (Papapanagos and Vickerman, 2000: 35), for the purposes of this project we define commuting as involving: the spatial movement of labour across borders which does not involve a change of residence. Therefore we are interested in citizens from country A who maintain their home in country A but work in country B whilst returning home to country A whilst not at work. This is a similar definition to the one used by Hitzelsberger et al. (2001) in their study of the extent of cross-border commuting in the EEA (European Economic Area). But there are problems of demarcation (e.g. transport workers such as a lorry driver) or where citizens from country A reside in country B but commute back to country A, as for example happens on the Netherlands/German border because of differential property prices. Nevertheless, in this exploratory study we have been guided by the above definition without being restricted by it. Aim of the project Specifically, the project had four overall objectives 1. to collect evidence and review existing information on factors which inhibit cross-border labour commuting in EU member states; 2. to gather, by means of small-scale feasibility studies, supplementary information on the perception of obstacles to seeking work in the neighbouring country; 3. to disseminate preliminary findings and generate hypotheses which will inform further comparative research; 4. to prepare a future comparative research project on this topic within the framework Programme vi which will involve more network members, including colleagues from accession countries. The six borders investigated were Finland/Sweden, Sweden/Denmark, Denmark/Germany Germany/Netherlands, the Netherlands/Belgium and Italy/France. These border regions were selected on the basis of a geographical spread and the diverse volume of flows of cross-border commuting to work. The research project focussed on identifying actual and, in particular, perceived obstacles and barriers to cross-border commuting which we expected to be different across the different borders selected. Stage 1 of the research involved the compilation of state-of-the-art surveys of ongoing and recent research in the area of cross-border commuting in the EU, and in six border regions in particular. Each of these border specific surveys were carried out by the relevant border team and made generally available as CROBOCOB working papers. These can be found as appendices to this report (7) and on the project website ( 7

8 Stage 2 of the project involved interviews with individuals in seven countries. A total of 102 interviews were conducted. The reports on these can be found as appendices (7) and on the project website ( 3. Scientific description of the project results and methodology The following provides a brief descriptive account of information regarding the border regions needed to put the subsequent analysis into context (3.1). The is followed by a review of possible factors which impinge on the propensity for cross-border commuting (3.2). The analysis (3.3) is divided into a summary of the main points taken from the country reports produced (see Annexes under 7) and a comparative account of major themes arising from within the interview data Extent and composition of cross-border commuting Overall, there is relatively little cross-border commuting to work within the EU. In the mid 1990s only some 0.2% of the EU labour force (600,000 workers) resided in one member state and worked in another. The pattern is one of great discrepancy between regions (de Falleur and Vandeville, 1995). This discrepancy makes the topic an ideal one for comparative study investigating the perceived and actual barriers which may inhibit cross border commuting. Applying Hitzelsberger et al s (2001) definition (see above), there were about 500,000 cross-border workers in 1999, which represented 1.4% of the total workforce in the border regions within the EEA. However, there seems to have been a significant increase (of 29%) in cross-border commuting between 1995 and Particularly the number of Germans commuting into the Netherlands rose significantly during the same period, annually by 30%. More than 80% of all commuters come from 4 countries (France, Germany, Belgium and Italy) and just over 70% of commuters travel to three countries to work: Switzerland (35%), Germany (20%) and Luxembourg (16%). 85% of all commuters can be found in the narrow region covered by the Benelux countries, the Rhine Valley and Switzerland. This is not surprising given the density of population in these regions and the longer history of cross-border cooperation. As far as the border regions in our project are concerned, based upon Hitzelsberger et al (2001) we have calculated the following proportions of cross-border commuters. Table 1:Cross-border commuting, selected border regions Cross-border region Total workers in region Number of commuters (year) % of commuters to workers in the region B/NL 3,412,000 22,940 (2000) 0.7 NL/D 4,331,000 33,107 (2000) 0.8 D/DK 325,970 2,480 (1999) 0.8 DK/S 2,345,200 3,000* (2001) 0.1 S/SF 224, (2000) 0.4 F/I/Monaco 2,535,000 1,007 (1999)** 0.04 Calculated from: Hitzelsberger et al., 2001, Table A2; * based on company surveys, expert talks, recent studies etc.; ** based on Hitzelsberger et al., 2001, Table A7 and partly based on company surveys, expert talks, recent studies etc. 8

9 Table 1 illustrates the low level of cross-border commuting generally. Even the relatively high level of commuting between Germany and the Netherlands, for example, remains below one per cent of workers in the region. The pattern of commuting between Denmark and Sweden is low as might be expected given the very recent nature of the bridge between the two countries. The proportion of the workforce commuting between Finland and Sweden, albeit in a sparsely populated area is low and the level between Italy and France is a twentieth of that in some other countries, which may only in part be because of topography and transport links. According to Hitzelsberger et al (2001) industrial sector workers are over-represented while private service sector professionals are under-represented amongst commuters, with hardly any public sector workers. One in three cross-border worker is female, while those with higher education degree represent the strongest growing group. Table 2 (below) suggest that not only is the proportion of the workforce commuting greater in some of the regions studied in this project but also that the significance of commuting as a European phenomenon is also linked to those regions. Table 2: Cross-border commuters in selected EURES regions; directions of commuting, 1999 EURES region Direction of commuting Number of commuters Scheldemond B to NL 2,060 Maas Rijn NL to B B to NL NL to B 1,760 10,760 1,330 D to NL NL to D Gronau-Enschede D to NL NL to D Rhein-Waal & Rhein-Maas- D to NL Nord NL to D EMS-DOLLART D to NL NL to D 5,869 8,664 2,517 2,527 7,355 4, EURAZUR TRANSALP Sønderjylland Øresund F to I I to F F to I I to F DK to D D to DK DK to S S to DK ,339 1, * 2,500* Tornedalen SF to S S to SF 600* 320* Source: Hitzelsberger et al., 2001, Table A7; * based on company surveys, expert talks, recent studies etc.. 9

10 3.2. Factors which influence mobility and immobility The importance of studying immobility, in a number of regions and through in-depth interviews has been pointed to by members of our network (see Janssen 2000; De Gijsel and Janssen, 2000). Indeed, while there are some studies which concentrate upon immobility (e.g. see Fischer/Martin/Staubhaar, 1997), investigations have concentrated on mobile workers rather than on those who do not commute to neighbouring countries. Overall, there is very little in-depth qualitative work being done to look into how factors of immobility are perceived by potential commuters. Border influences can be divided into two broad groups (van der Velde, 2000). First: objective physical and institutional constraints; and second: the subjective values and meanings given to cross-border activities, and subjective perceptions of the differences on both sides of the border. Our project sought to tap into both of these by exploring the border as 'a mental map'. We aimed to identify the key features on the border map of respondents which contribute to immobility. One innovative aspect of our project was its comparative focus on factors which might influence immobility. Much of the previous research in the area has tended to be focused on a single cross-border region (sometimes two), has concentrated on labour migration (workers changing residence) and tended to be quantitative in nature, approaching the topic from an economic perspective seeking explanations of cross border commuting in terms of utility maximisation (e.g. Papapanagos and Vickerman, 2000). Hansen and Nahrstedt (2000) sampled two thousand three hundred cross-border commuters on the Denmark-Germany border and identify a number of factors which impinge upon commuting. They point to standard economic explanations, 'in theory people tend to commute from countries with low wages and high unemployment to countries with high wages, low unemployment and prosperous industrial structures' (Hansen and Nahrstedt, 2000: 71). Werner (2001) points to similar economic pull and push factors which might facilitate migration of commuting. Better job prospects and wage rates elsewhere are potentially examples of the former, while unemployment and low income would strengthen push factors. Werner (2001) argues that workers in border regions respond quickly to changes in the neighbouring country and that some barriers which are familiar to migrants, are of lesser importance for commuters. For example, there is no need to move house, the partner can keep his or her job, children remain in the same school and language problems are less significant than for migrant workers. Unlike migration then, cross-border commuting might become more significant in scale because regional economic areas near borders will grow together even more in the context of a more economically integrated European Union. But which factors influence the propensity to commute across borders? In addition to economic aspects (wages, unemployment; benefits), other factors can be identified which might influence the willingness to commute to work in a neighbouring country, such as socio-demographic factors (age - the young are more likely to be mobile; education - whilst the evidence is mixed the better qualified appear to be more mobile; marital status - family responsibilities may reduce mobility but may increase the likelihood of cross-border commuting), geographic factors such as cross-border transport and communication links, 10

11 information, particularly labour market or housing market information, personal factors such as professional contacts, experience of cross-border commuting, existence of social capital, and personal and family ties; cultural factors such as language or lifestyle, legal factors (such as the recognition of qualifications, social protection systems or access to public sector jobs). Existing research touches on some of these issues. Tassinopoulos et. al. (1998, 1999), for example, present four arguments for immobility: risk aversion; discrimination; social protection systems; and legal barriers. Others (Rietveld, 1993) have identified lack of information and personal information networks which relate to the wider cultural links and which may contribute to immobility (de Falleur and Vandeville, 1995). In relation to social protection, national differences in social security systems and fiscal systems (van der Veen, 1993) may create problems but may also be an incentive (Hansen and Nahrstedt, 2000; see also Bacher et al, 1995; Bode et al, 1994). For example, in the Danish-German border region cross border commuters pay contributions and income tax at their place of work whilst their rights to benefits and services relate to residence (Hansen and Nahrstedt, 2000). The recent data by Hitzelsberger et al. (2001) show a clear positive correlation between the level of commuting and population density. From a reverse perspective, one could ask not so much about the barriers to mobility but the advantages of immobility. Based on the seminal work of Straubhaar (1988), Tassinopoulos and Werner (1999) and Fischer et al. (2000) emphasise location-specific or insider advantages which would be lost if people were to move away (lost sunk costs, i.e. costs that are tied to a specific location which are lost in case of mobility). The value of immobility includes both work-oriented advantages and leisure-oriented advantages. The former relate to firm-specific non-transferable knowledge (generating higher firm-specific compensation), space-specific advantages (e.g. expertise in regional preferences, habits of clients, locational production technology and other forms of insider knowledge) or society-specific advantages (social relations and political activities linked to local residence). Leisure-oriented advantages include societyspecific aspects (social networks, social inclusion and integration, political participation) and space-specific advantages (e.g. knowledge of cultural events, housing market etc). The accumulation of these advantages takes time and effort which implies that, once acquired, staying immobile has its own value (Tassinopoulos and Werner, 1999: 12). Other reasons for immobility include risk aversion (it is more difficult to assess risks in a foreign country) and possible discrimination against foreign workers, particularly when jobs are scarce and social security systems. Even if benefits are transferable and social security differences across Europe are diminishing, the perception of possible losses of social rights and benefits can act as a deterrent to commute. The same applies to legal barriers. While Tassinopoulos and Werner (1999) argue that these should be of less importance for inter-european migration, individually people might perceive legal obstacles as difficult to overcome. Finally, there is the time aspect. There might be a value of waiting (i.e. not seeking work across the border yet). The situation in the location of residence might improve, 11

12 the one across the border deteriorate and more information can be gained before work is being applied for. Also, the consideration of commuting across borders might be based on a long-term perspective or for short-term reasons, which might have different implications for the assessment of benefits and risks of commuting. Based on information from EURES co-ordinators and partners (for EURES, see point 4 below), EURES (European Commission 2002) commissioned a report which provides a typology of obstacles of political, economic, geographic and cultural nature in most EURES regions. For example, for Euregio Gronau/Enschede, it lists and details the ways in which family allowances, health care, patient costs and sickness benefits, social insurance, unemployment benefits, pensions and taxes can be considered to be objective obstacles for potential commuters. Other partnerships, such as EURAZUR, list and discuss similar aspects of social security in addition to employment contracts, bureaucracy, language problems, job postings and vocational training. This type of information was highly relevant for the empirical part of our project Analysis This study concentrated upon six borders. These are Finland/Sweden, Sweden/Denmark, Denmark/Germany, Germany/the Netherlands, the Netherlands/Belgium and Italy/France. These borders were chosen on the basis of a geographical spread and the diverse volume of flows of cross-border commuting to work. In four cases we investigated two borders from the same country (Sweden, Denmark, Netherlands, and Germany). In one case we looked at a very sparsely populated border area Finland/Sweden. In one case the border is very heavily populated (Netherlands/Belgium/Germany). This region also includes a border without a language barrier (Netherlands/Flanders). In one case there is a border where recent transport infrastructure has increased the potential for commuting (Sweden/Denmark). The similarities and differences between these borders are discussed in more detail below. Research teams collected data about the border regions according to a template agreed upon at workshop I, held in Stirling in January Subsequently, in depth interviews were conducted in four border regions (both sides of each border; 102 interviews in total) with people who were not commuting but who might be in a position to do so. The following summarises the national labour market conditions; the extent of cross border commuting; barriers to commuting; and policies in the border region for each of the border regions as reported by the research teams (for more detailed information, see Annexes, point 7 below). It then proceeds to the comparative analysis of both the national reports and, subsequently, the interviews as a whole (for border specific interview analyses, see Annexes, point 7). Finland/Sweden Labour market conditions Finland experienced a higher rate of unemployment than the EU average in This has been accompanied by labour market shortages in southern Finland and a higher rate of unemployment in the north and east of the country. In particular following the increase in unemployment which took place in the early 1990s, a skills mismatch has developed alongside rising levels of long-term unemployment and a policy concern with 12

13 the regional differences which have developed especially where there is an ageing of the population. This has lead to a concentration upon developing the skills base. The changing demographic structure in some parts of the country has lead to a concern with avoiding early retirement, encouraging immigration and raising the activity rate of the working age population. In common with Sweden, Finland has high levels of employment in the public sector, particularly amongst women. Sweden has pursued a strategy of full employment and economic growth. This is achieved through channelling work to the unemployed and labour to employers. The aim is to combat bottlenecks in the labour market through skills enhancement and activation initiatives for those who have difficulty getting back into the regular labour market. In common with Finland there are regional disparities in levels of employment within the context of a relatively high rate of labour market participation. In some regions underemployment is a problem. In Finland a larger proportion of the population do not have any further education qualification than in Sweden. This in part is a reflection of a different age structure in the population, the older being less likely to have FE qualifications. The same age gap in qualification exists in Sweden but there is not such an extreme gap. There are a number of similarities in the Finnish and Swedish labour market which are worth noting: regional disparities in employment and skills (particularly north/south) high levels of public sector employment rapid increase in unemployment in the early 1990s the ageing of the population and consequent demographic change skills shortages accompanying unemployment. Cross border commuting Migration within both Finland and Sweden tends to be internal. There are however two contrasting border regions. In the north there is the sparsely populated area of Lapland where the Swedish and Finnish border is marked by a river. Further south there are the Finnish Alund Islands which lie halfway between Sweden and Finland. The Alund Islands have a very low level of unemployment with a tradition of outmigration to Finland as an alternative to unemployment. In the north unemployment is higher than the national average on both sides of the border. In the north of Finland there are high levels of unemployment accompanied by low levels of education and high levels of public sector employment as well as employment in forestry, mining and construction. In general communication links tend to be internal to the two countries with poor eastwest transport links. Barriers to commuting Five barriers to commuting in the chosen region have been identified: Sweden is not in the Euro zone whereas Finland is which leads to currency fluctuations effecting take home pay; The high levels of public sector employment restricts commuting as public sector jobs may not be available; Language is a barrier in that there are three languages in use Swedish (although there is a Swedish speaking population living in northern Finland), Sami spoken on both sides of the border by Laps, and Finnish which is not understood by Swedes; Limited bridges across the border river; 13

14 There are differences in the vocational training systems in the two countries Border region This border region is a contrast to the other border regions in this study in that it is very sparsely populated. The main centres of population on the border are small. In Finland there is only one municipality exceeding 10,000 people and in Sweden the largest has 20,000 people. Policies have been developed to ensure that the social benefits are harmonised across Nordic countries allowing for migration across borders. In the study area the Tornedalen partnership a EURES project on the Lapland/Norbotten border has sought to develop cross border structures and labour market information with the aim of increasing the vacancies available for those in the border region. An INTERREG III programme provides a website of employment agencies for those interested in the labour market in the area. Whilst there are higher rates of unemployment on the Finnish side of the border, commuting numbers are small. In 1999, 600 people commuted from Finland to Sweden and 320 commuted from Sweden to Finland. Germany/the Netherlands/Belgium Labour market conditions These borders have experienced contrasting levels of unemployment. In the case of Belgium and Germany, unemployment exceeded the average of other similar countries in the late 1990s and has remained high. In addition Belgium has one of the lowest rates of employment and highest rates of long term unemployment in the OECD. In the case of the Netherlands, unemployment peaked in the 1980s and fell in the 1990s. In terms of their labour markets the countries are an obvious contrast: the size of Germany with its remaining very high levels of unemployment in the East of the country; the language divide in the small country of Belgium; and the tight labour market in the Netherlands. Flanders which was the study area in Belgium has a border both with Germany and the Netherlands, has a shared language with the Netherlands and is one of the areas in Belgium with the lowest rate of unemployment. Labour market policies in the three countries have some similarities concentrating upon increasing flexibility, reducing labour costs and reintegration of the unemployed albeit starting from different conditions and with different institutional structures for labour market policy. In common with other countries mobility is higher within national borders amongst the better qualified workers. Cross border commuting On the Dutch German border there are five Euroregions with higher than average unemployment on both sides of the border towards the north and lower than average in the southern regions. In all parts the Dutch rate is lower than the German. If unemployment is a major factor in cross border commuting then commuting from Germany to the Netherlands would be expected. Barriers to commuting In contrast to some of the other border regions examined the infrastructure exists for commuters with a good traffic infrastructure and settlements which provide the potential for daily commuting. There is also a long tradition of cross-border cooperation. There is evidence that in this border area of residents moving across the borders to commute back because of differential property prices. Public debate on Dutch-German cross 14

15 border commuting has been dominated recently by concern with the low levels against the background of high rates of unemployment on the German side of the Dutch/German border. In the case of the Dutch/Belgian border there is evidence that practical problems were a major problem for those who did commute although new agreements between the countries have attempted recently to deal with some of these for example in the field of taxation. Border region Mobility in these border regions has always been relatively high. The interviews on the German-Dutch border were carried out in the Euroregions of Rhine-Waal/Rhine-Maas- North which are in the centre of the border. In Rhine-Waal, the Dutch side of the border has two cities whilst the German side is more rural whereas the picture is reversed further south. In addition commuting has increased from Germany to the Netherlands in Rhine-Waal. In 1998 the ratio and absolute numbers of cross-border commuters increased from north to south with a higher proportion of the Dutch workforce commuting into Germany than vice versa. Since 1997 the numbers of companies in which Germans are working in across the border has almost doubled. Whilst Dutch companies have active recruitment policies in Germany to deal with labour shortages, this is not the case for German companies. Despite the greater extent of cross-border commuting than on many other borders, cross-border commuting remains low compared to commuting within the countries. Italy/France Labour market conditions The French labour market has both created jobs and reduced unemployment in recent years, although the unemployment rate in 2001 was still higher than the EU average. However, whilst unemployment in the border region with Italy has been declining, it remains along with other French Mediterranean regions with unemployment higher than the French average. Whilst Italy has also experienced falling levels of unemployment recently, the highest rates of unemployment are experienced in the south, with much lower levels in the border regions. French unemployment policy has concentrated upon reducing working hours to redistribute employment, and more recently job creation and reducing labour costs. In Italy policy has concentrated upon trying to deal with some of the rigidities in what has traditionally been a highly regulated labour market and through the creation of employment incentives for particular groups of workers most likely to experience unemployment (the young, the long-term unemployed etc.). There is a contrast in the French and Italian systems of taxation of income with income tax being levied by employers in Italy and through self-assessment in France. Whilst there are agreements which deal with the position of cross-border commuters, these different systems are worth noting particularly in how they may effect marginal cross-border commuters. In the border region, the service sector places a dominant role in the local economy on both sides of the border. Cross border commuting Cross border working between Italy and France is relatively low. Most Italian workers travelling across the border travel into Monaco. The numbers of French crossing the 15

16 border into Italy are negligible. The evidence suggests that the characteristics of those who do commute across this border are quite different in the different directions. Italian cross-border workers tend to be medium low skilled whilst French workers are more likely to be highly skilled professionals. Barriers to commuting EURES suggests that the main barriers to commuting relate to social security systems and work practices. In addition the recognition of education and training qualifications and language skills act as barriers. In particular French workers lack information about job opportunities in Italy and face language barriers. French workers also have to go through an administrative complex procedure to gain a special permit to work in Italy. Border region The border region has strong historical cross border links and a good transport infrastructure along the coast. The service sector is the predominant employer on both sides of the border centred on tourism. Whilst the levels of unemployment on the Italian side of the border are lower than on the French side these overall lower levels of unemployment reflect in part the availability of seasonal jobs in the tourism industry. Border commuting from Italy has decreased in recent years. Those who do commute tend to be younger, male and lower skilled. There is limited information on cross-border workers from France. The EURES partnership aims in particular to address the low volume of cross-border flows. Germany/Denmark Labour market conditions Denmark is characterised by a full employment policy based upon active labour market policy to encourage the unemployed back into work. The aims of employment policy are to encourage job creation and increase labour market participation. Unemployment peaked in the mid-1990s and today Denmark has one of the highest rates of labour market participation and lowest unemployment amongst EU countries. On the Danish- German border unemployment in Germany is higher than in Denmark Cross border commuting In the mid-1990s a cross-border Regional Council was established to discuss and promote cross-border cooperation. At the same time two labour market initiatives were launched which advise people on cross-border commuting and raise awareness of the labour market in the region. Evidence on cross-border commuting is derived from a study in the late 1990s which indicated that commuting had remained fairly stable from the end of the 1980s to the mid-1990s. Slightly more people commuted from Denmark to Germany than vice versa. Cross-border commuting does not seem to have an impact on the overall labour market, although cross-border commuting is concentrated on particular parts of the border. Most of the commuters have a higher education qualification and are between 30 and 50. The evidence suggests that the largest group of commuters are those who have taken advantage of property price differentials and have moved their residence from Germany to Denmark but kept their employment in Germany. Overall however almost equal numbers commute each way 16

17 Barriers to commuting There are a number of differences in the organisation of social security which may impact upon labour mobility. In addition the systems of tax deduction and tax allowances and the tax unit of assessment may create problems for cross-border commuters. According to EURES there are linguistic and cultural barriers to crossborder commuting which are reinforced by a lack of knowledge and information about laws and regulations as they affect commuters. Border region The main industries in the border region are manufacturing, agriculture and tourism. The public sector also employs a high proportion of workers on both sides of the border which is of significance where there are restrictions on the employment of non-nationals in public sector jobs. The region contains nearly 700,000 people with the majority in Germany. More than two thirds live in small towns or villages and the region is sparsely populated compared to the rest of Denmark or Germany. Denmark/Sweden Border region The Øresund region has benefited from the opening of the bridge across the strait in 2000 which links Copenhagen and Malmö. Approximately 3.5 million people live in the region, two-thirds on the Danish side. The region includes both rural areas and a large metropolitan area. The bridge has lead to an increase in cross-border activities. The majority of the workforce is employed in the service sectors. There are considerable similarities across the border in terms of the structure of employment, levels of skills and wages. However there is a disparity in employment growth across the border with higher levels of employment, lower unemployment on the Danish side, although the disparity has been decreasing as the Swedish side of the border has experienced high rates of growth. This has lead to labour shortages in similar sectors on both sides of the border. Institutional mechanisms involving both the public and private sectors have been developed to promote cross border economic cooperation with the aim of integrating the labour market in the region and formalising cooperation. Cross border commuting Commuting from Sweden to Denmark is much higher than in the reverse direction, although the numbers remain small considering the extent of commuting within the region which takes place within each country. The commuting from Sweden to Denmark may be accounted for by lower housing costs in Sweden and more employment opportunities with higher wages and shorter working hours in Denmark. Cross-border commuters tend to work in larger companies. Barriers to commuting There are important differences in how the social security and tax systems work in the two countries which may have an impact upon cross-border commuting. This particularly arises with the use of tax deductions and allowances which may be based upon place of residence rather than place of employment. EURES indicates that 80% of enquiries concerning working between the two countries relate to tax. The rules governing social security benefits and contributions in the two countries are also 17

18 complex and involve an interaction between place of residence and place of employment. Recent studies of attitudes towards cross-border commuting in the region indicate that travel time and cost were seen as major barriers along with uncertainty about tax and social security arrangements and, to an extent, language. Commonalties and contrasts in the border regions A reading of the accounts of the national labour market conditions, the extent of cross border commuting, the barriers to commuting and policies in the border region provided by research teams points to some commonalities and contrasts in the national context within which the interviews took place. In terms of commonalities, the country reports point to features which are shared across many EU countries such as the impact of demographic structure on labour market supply and demand. In many countries the most disadvantaged in the labour market are those young people without skills or qualifications as well as older workers with either old or few skills. All of the countries have experienced (although within a different cycle) rising unemployment during the 1990s and the development of an increasing emphasis upon activation policies for the unemployed combined with attempts at welfare state reform and change. In many of the countries, public sector employment is an important part of total employment which has a particular impact upon cross-border commuting where access to some public sector jobs is restricted for non-nationals. In terms of contrasts the border regions that have been examined present difference in terms of geography, population, transport links, institutional cooperation and language. The borders include sparsely populated areas Sweden/Finland in contrast to the relatively densely populated borders of the Netherlands/Germany/Belgium or the France/Italy border where the population is concentrated on the southern part of the border. The Finnish/Swedish border also represents a contrast as a peripheral area in both national and EU terms and the centrality of the Netherlands/Belgium/German border. On the other hand there are borders such as the German/Danish or the French/Italian which whilst peripheral in national terms are important transport routes in EU terms. The borders also have variable communication infrastructure. Whilst in most countries internal communications are better than cross-border communications in some cases major communication routes cross the border for example between France and Italy and Sweden and Denmark. Whilst all the borders have some institutional form of cross-border communication, these vary both in their extent and in the length of time that they have been running for. Again the German/Dutch border has a long history of cross-border cooperation compared to the more recent institutional forms developed on the Swedish/Finnish or French/Italian border. The extent of the participation of social partners and the degree of formalised institutional cooperation also varies with strong models on the Danish/Swedish border or the Danish/German border compared to that on the French/Italian border. The borders also vary according to the labour market conditions in the border regions. Here there are examples of high unemployment on both sides of the border e.g. Finland/Sweden; relatively low unemployment on both sides e.g. Sweden/Denmark and 18