- Introduction
The UK government is committed to ‘control’ of EU immigration after Brexit. However, the movement of workers can be ‘controlled’ in different ways, with a distinction between quantitative (e.g. quotas) and qualitative (e.g. standards requirements) controls. ‘Quantitative’ controls, although seemingly straightforward, may prove problematic for a number of reasons:
- As the government plans to keep tourist and business travel between the UK and EU visa-free, and the Irish border open, even the strictest controls would not reduce the pool of people available for undeclared work;
- UK employers are demanding a degree of post-Brexit access to the EU labour pool, and strict numerical controls are unlikely to provide the necessary flexibility in a volatile economy;
- In the negotiations with the EU, there is clear trade-off between limits to migration and access to the single market, especially in services, as evident from the Swiss experience and from multiple declarations of EU-27 leaders;
- Numerical restrictions address yesterday’s problems: inflows of immigrants from the EU new member states have already declined by 50% since the referendum and are expected to fall further for demographic (very low birth rates in Central Eastern Europe since the 1990s) and economic reasons (rapidly falling unemployment and wage gap from the UK).
‘Qualitative’ controls are thus an important option, with labour market regulation affecting the immigration issue in two ways. First, it can reduce local populations’ concerns with ‘social dumping’ and negative effects on wages, employment and working conditions. Second, it can reduce employers’ demand for foreign labour by narrowing/removing the wage differential between the local population and migrants, and/or by adding extra costs such as controls and permits. Further, the involvement of social partners and/or technocratic agencies may potentially ‘depoliticise’ the issue. While improving standards might in theory act as an additional ‘pull’ factor, in reality, intra-EU movement of workers is largely demand-driven and unaffected by labour standards consideration (as proved by larger flows towards the UK, Ireland and Southern Europe than towards Sweden or Denmark).
This paper analyses these different approaches by looking at the recent experiences of Canada, Norway and Switzerland, often mentioned as alternative ‘models’ for future UK-EU relations. All three have larger numbers of foreign-born workers, and larger recent inflows in relation to their population, than the UK. Further, all three operate different combinations of quantitative and qualitative controls, thus providing instructive examples. This analysis is based on documentary analysis, secondary literature and interviews with social partners, NGOs and policy makers in the three countries, carried out in May-June 2017. The review of the policies of the three countries ends with recommendations, with the warning that no country provides a ready ‘model’ that can be directly transposed to the UK.
- Norway
Norway is a member of the European Economic Area (EEA) and is thus covered by the principle of free movement of workers. The EEA Treaty foresees the possibility of a unilateral ‘emergency brake,’ but Norway has never considered using it, in part because of the potential risk of EU countermeasures. Instead, Norway has adapted its labour market regulations to the risk of ‘social dumping’ through migrant labour. This started in 1993 when a law on the general extension of collective agreements (General Application Act 58/1993) was introduced, allowing for the extension of provisions of sectoral collective agreements to all employees in the sector. The Act constituted a significant innovation to the extant model of bilateral, rather than state, regulation of the labour market.
However, the Act remained ‘dormant’ until the 2004 EU enlargement, when the country received a large influx of immigrants. In response, trade unions, and sections of the employers, activated the General Application Act in sectors with significant presence of foreign workers. By 2017, sectors covered by binding agreements include construction, shipbuilding, off shore oil industry, agriculture, cleaning, electricians, fish processing, and road freight and passenger transport.
The employer side in Norway has been divided on the issue, particularly between different sectors, with the NHO umbrella organisation having a mediating role. In sectors shielded from international competition, such as construction (where agreement extension occurred in 2007, supplemented by further measures such as joint liability for sub-contractors and compulsory ‘ID cards’ proving legal employment status), employer associations supported the legal extension of agreements in order to maintain a common floor for fair competition. In more exposed sectors, such as shipbuilding, there has been strong resistance and a protracted (and as yet unresolved) legal dispute in the EFTA/Norwegian Supreme Courts over the treatment of posted workers. Extension is supported by an inspection regime, with the Ministry of Labour collaborating in bilateral agreements with labour inspectorates of sending countries in Central Eastern Europe.
Agreement extension has ensured that, since the early 2000s, average real wages have continued to increase, including in sectors with high immigration. ID cards in construction and cleaning, as well as chain liability, have helped enforcement and kept self-employment low at 7%. Strains on labour conditions (e.g. workers being employed at a lower level than their qualification, or not benefiting from ‘wage drift’, i.e. higher company wages above the minimum sectoral levels) have not been eliminated, but multiple evaluations have shown that extension has allowed the introduction of an important, if imperfect, floor in core labour standards. Further open issues associated with migrant labour – agency workers’ employment conditions, zero-hour contracts, public-sector outsourcing are also being tackled through proposals for further labour regulation, rather than limits to migration, thanks to a cross-party consensus on the use of standards to combat social dumping.
Politically, as confirmed by ongoing surveys, stricter regulations appear to have kept popular concerns over immigration at a relatively low and constant level, containing demands for numerical controls of free movement.
Overall, the use of innovative labour market regulations, combined with social partners’ involvement and legal intervention, has allowed Norway to face an unprecedented influx of migrants, both without major social disruption and the need for quantitative limits on free movement. It is less clear if they may have also reduced employers’ demand for labour: arrivals have fallen in recent years but also as a result of a declining value of the Norwegian Krone.
- Switzerland
Switzerland rejected EEA membership in 1992 but since 1999 has signed a series of bilateral agreements with the EU, broadly similar to EEA arrangements. Free movement was therefore introduced in 2004, following a 1999 Bilateral Agreement on free movement, which led to the replacement of the previous system of contingents and work permits. Immigration has long been a sensitive political issue in the country. Switzerland has the second-highest share of foreign residents among EU/EFTA countries, although the majority of EU residents are not from Central Eastern Europe, but from other western European countries. While in the late 1990s Switzerland was a net emigration country, since the introduction of free movement larger inflows have expanded the population from 7.5m to 8.1m within a decade.
After a 2009 referendum which approved free movement extension to Romania and Bulgaria, a vote in 2014 narrowly resulted in the introduction of quotas on EU immigration, to be introduced within three years. However, as strict quotas would violate the bilateral agreements, with potential economic damage for a country dependent on EU trade, the Swiss Parliament instead introduced, in December 2016, the prioritising of Switzerland-registered job-seekers in job recruitment. This solution has been acceptable to both the EU and, according to opinion polls, the majority of the population.
This measure will work in conjunction with the extant system of ‘flanking measures’ introduced since 2000 with the aim to prevent social dumping. These measures take the form of generalised labour market controls, binding sectoral collective agreements, as well as tripartite setting of minimum standards in sectors without sectoral agreements. The number of labour inspectors has continued to increase as has the stringency of sanctions. By law, at least 2% of employers (3% in high risk sectors) and 50% of posted workers and foreign self-employed must be inspected every year. In sectors with binding collective agreements, inspections are the responsibility of bipartite commissions of employers and trade unions.
Employers support the system as a means to ensure fair competition, although they would prefer fewer, more targeted controls.
Regarding binding collective agreements, the 1999 flanking measures introduced a simplified procedure to declare the binding extension of sectoral collective agreements, at canton or national level, in sectors with evidence of social dumping. This increased collective bargaining coverage from 45% to 51%, the number of extended agreements increased from 31 in 2001 to 73 in 2014 (including critical sectors such as construction, cleaning and temporary agencies), and the number of employees covered by them from 400,000 to 1m.
In sectors without binding collective agreements but with evidence of social dumping, the measures allow for the introduction of cantonal or national sectoral legally-binding ‘normal employment contracts’, specifying minimum wages and standards. These are introduced by tripartite commissions, including social partners and cantonal governments. ‘Normal employment contracts’ tend to be less generous than collective agreements, and given the resistance of some cantonal governments, they have not been used unevenly across the country. The relative weakness of such contracts is significant as in recent years, the share of foreign workers has increased faster in sectors without binding collective agreements.
In terms of outcomes, regulations have kept the wage distribution constant and the rate of self-employment at 14%. Some problems do remain. Foreign employees tend to be classified at the lowest end of the wage scales; a degree of pay undercutting remains. Rising immigration in construction has resulted in a sharply growing share of employees classified as non-qualified, suggesting that many are being employed at a lower level than their actual qualification.
However, Swiss free movement policies are unique for the extension of labour market restrictions and controls. While they have not prevented anti-EU votes, they remain the default response by all main actors to popular demands. Overall, the measures have contained wage dumping in critical sectors such as construction, limited the negative aspects of self-employment and posted workers and, together with the recent introduction of residents’ priority (which is too recent to be evaluated), have maintained sufficient consensus across industry, politics and population on maintaining free movement. Labour market measures appear to have strongly affected public opinion historically: after the No to the EEA in 1992, the introduction of the ‘flanking measures’ in the early 2000s contributed to a strong Yes to extending free movement in 2009. And whilst more liberal EU policies in the early 2010s led to a Yes to quotas in 2014, the introduction of residents’ priority satisfied the large majority in 2017. As in Norway, it difficult to reach a definite conclusion on the effects on employer demand, but the relatively low recruitment of Central Eastern Europeans suggests that an effective floor to wage competition is in place.
- Canada
In 2016, Canada signed the Comprehensive Economic and Trade Agreement (CETA) with the EU, which does not include free movement of people. Canada’s migration model, and specifically its point system, had been presented as a model during the UK’s EU Referendum campaign. While such system is often presented as managing immigration according to the needs of the economy, the historical analysis of immigration inflows and unemployment trends in Canada reveals a low and declining adaptation of migrant inflows to job vacancies/unemployment trends, in part because the point system is intended to address long-term demographic and human capital needs, rather than short-term economic ones.
As the point system is oriented towards highly qualified individuals, Canada has had to open other channels to meet the need for labour in sectors characterised by lower wages, low skills/low formal educational requirements. This has occurred through the Temporary Foreign Workers Program (TFWP), originating at provincial level but expanded at federal level in 2002 (Canada has a complex distribution of responsibilities between federal and provincial levels). The TFWP allows work permits for up to two years subject to Labour Market Impact Assessments (LMIA). Unlike in Norway and Switzerland, LMIAs are largely employer-led with little trade union involvement.
Following the TFWP’s expansion, the 2004 conservative government further extended it through the International Mobility Program (IMP), which provides temporary work permits (open or employer-specific) of up to four years’ duration, for nationals of certain countries (e.g. NAFTA) without the need for LMIA. On permit expiration, workers must leave the country if permanent residence is denied, with the success rate for such applications being generally low. The number of temporary foreign workers increased sharply during the 2000s, while the share of high-skilled workers among them declined from 67% to 40%.
The TFWP has been widely criticised for the ‘conditionality’ of employer-tied work permits, which makes foreign workers virtually ‘captive’ of their employers. Changes of employer can only happen within the same sector and with the agreement of the first employer. Employers and government argue that without conditionality, foreign workers would quickly move away from low-paying industries.
Media attention on cases of migrant exploitation and substitution of Canadian workers in a number of sectors swiftly led to changes in 2013-14. In the run-up to the 2015 federal elections, the conservative government introduced draconian restrictions on the use of temporary foreign workers, including strict quotas (10% cap in the workforce) and much higher fees (from 100 to 1,000 CAD). These changes were strongly opposed by business but also met criticism from trade unions and migrant NGOs. These openly punitive restrictions did not secure the Conservatives’ electoral victory, and the subsequent liberal government withdrew some of the limits and introduced some worker-friendly measures, e.g. the removal of foreign domestic carers’ obligation to live at the domicile of their employer/sponsor. Moreover, concessions were made towards more permanent routes to immigration (e.g. the ‘Atlantic Immigration Pilot Program’) and more sector rather than employer-based work permits.
Even if Canada has a liberal labour market with decentralised collective bargaining, there are examples of co-ordinated social dialogue responses. In the meat-processing industry, collaboration between unions and large employers allowed combining continued support for immigration with better enforcement of standards across the industry.
More effort has been made in recent years to improve labour standards enforcement, although labour inspection efficacy varies by sector and province. Manitoba is the province with strictest regulation, ensuing from the 2008 Worker Recruitment and Protection Act. This introduced stricter pre-recruitment labour market assessments to avoid substitution of nationals, a compulsory register of all employers recruiting foreign workers, proactive enforcement mechanisms and the largest provincial Permanent Nominee Program. Other provinces have since followed some of these policies (Ontario, Québec, Nova Scotia, Alberta and Saskatchewan).
In sum, the Canadian debate over immigration differs from that in Europe, as the country identifies as a country of immigration. However, despite its liberal employment regime, concerns with labour standards have gained prominence. In particular, temporary/seasonal work permits - the default policies in the mid-2000s - have come under increasing criticism from all political parties, and have been scaled down since 2014. However, this damaged business without improving conditions of employment. Several provinces are now moving towards more permanent permits and stricter labour standard enforcement.
- Discussion and recommendations
Geographic and institutional differences complicate direct policy transpositions. However, the analysis shows that in all three countries, the policy focus has gradually shifted from quantitative controls and conditional work permits to greater attention to migrants’ employment conditions.
The experiences of pre-2000 Switzerland and post-2000 Canada caution against temporary work permits, which tend to maximise foreign workers’ vulnerability. By contrast, in Norway, Switzerland and some Canadian provinces, better enforcement of labour standards has been instrumental in reducing popular concerns with immigration and, possibly, limiting employer demand for low-pay migrant labour. In several cases, employers also report having benefitted from better labour standards for immigrants, through improved retention and fairer conditions of competition.
The ‘liberal market economy’ of the UK may problematise the imitation of the Norwegian and Swiss co-ordinated solutions, but some British co-ordinated traditions (wage councils, collective bargaining, health and safety representatives) and the widespread recognition of the need for change after Brexit, allow for sanguinity on the feasibility of some specific measures. In particular, the following measures would be compatible with extensive access to the single market and, combined with an ‘emergency brake’, provide the effective ‘control’ of migration that people demand without damaging industry and increasing risks of exploitation:
• legal extension of collective agreements in sectors with multi-employer bargaining
• binding sectoral/regional minimum wages and minimum core conditions in sectors where collective bargaining is weak
• joint liability across the subcontracting chain
• compulsory registration and professional ID cards for workers in sectors with risks of undeclared work and bogus self-employment
• more extensive controls of employment conditions, involving industry and trade unions
• stricter regulations of posted workers, with obligation to pay the going wage rates
• local labour market checks to ensure resident recruitment priority in high-unemployment areas.
6 November 2017
