The impacts of migrants and migration into Scotland

3. Economic impact of migration

Key points

• The fiscal and economic impacts of migration are complex and many separate strands need to be analysed in the process of assessment.
• The topic has received a good deal of academic attention in recent years. However, the majority of the evidence is at the level of the UK, rather than Scotland.

Introduction

An accurate understanding of the fiscal and economic impacts of migration is essential to inform public debate. However, measurement is a complex task, and many separate strands need to be analysed in the process of assessment. These strands include migrant employment rates, the extent to which they pay taxes and receive government benefits, and their contribution to an economy's capacity for innovation ( OECD Policy Brief, 2014:1).

The body of evidence on the impact of migration on the UK's economy is large and growing. The two areas that attract the most attention are effects on the labour market and public finances (Lisenkova and Sanchez-Martinez, 2016:4). The sections below consider the evidence relating to each of these areas.

The Rolfe and Metcalf review noted that 'research on the macroeconomic effects of migration is UK-wide. Studies which focus on the economic impact of migration into Scotland address concerns at demographic change and the tightening of the labour market which is projected to follow population decline. Research has identified in-migration as the demographic variable with the biggest economic impact' (2009:3.3).

In general, this assessment of the evidence base still holds. The continuing UK-wide focus of evidence relating to migration and the economy might be expected, given that neither policy nor policy levers relating to migration are devolved to Scotland. Rolfe and Metcalf also identified a continuing preoccupation with demographic considerations as the way in which the migration narrative is framed in Scotland, rather than labour market needs, or the benefits of attracting human capital, for example. Again, this has not changed since 2009.

Impact of migration on the labour market

Key points

• The majority of migrants are young and economically active. Many are highly qualified, although qualitative research has indicated that there may be difficulties comparing qualifications across countries.
• Several studies have found a 'U-shaped pattern' of wage distribution among migrants, with employment concentrated at the top and bottom of the occupational distribution curve.
• From the perspective of employers, migrants are a flexible supply of labour to be drawn on when the economy is strong, or when labour demand exceeds local supply. However, agriculture is one area where overseas workers form an important part of the workforce, regardless of economic conditions.
• Migration does not appear to have had statistically significant impacts on the average wages and employment opportunities of the UK-born population in periods when the economy is strong, although there is some evidence of labour market displacement when the economy is in recession.
• The available evidence indicates that any adverse wage effects of migration are likely to be greatest for resident workers who are themselves migrants .

Migration can have many economic benefits. The international evidence indicates that skilled migrants can boost innovation, stimulate economic growth and encourage the local labour force to invest in training to take on and specialise in jobs in which the nation or region has a comparative advantage (Bell et al, 2014:311).

A central question for any cost-benefit analysis is whether migrants compete with workers in the host country, or whether they are complementary to them (Springford, 2013). If they are complementary, migrants will make the host population more productive, by doing work that others do not want to do, or do not have the skills for, or by introducing new ideas or technology. In practice, inevitably some workers will lose out. 'But if immigrants are on average complementary, it makes economic sense to let them in, as it will raise the productivity, and thus the average income of the host population' (2013:2).

In an examination of the economic and demographic consequences of large-scale migration for the UK, Rowthorn (2014) concludes that large scale migration can help to rejuvenate an ageing UK population by importing a large number of young migrants. Providing migrants are able to find jobs without displacing people who are UK-born, and providing they are sufficiently productive and well-paid, ' this will increase GDP per capita and generate a fiscal surplus for the government.' However, if migrants fail to get jobs, end up in low-skilled jobs, or displace UK-born workers, ' large-scale immigration will have a negative impact on GDP per capita and on government finances' (2014:66).

Key issues to consider, therefore, are the skills that migrants bring to the country; the type of employment they obtain and how they are viewed by employers; and the impact of migration on the wages and employment of UK-born workers. In the sections below, the evidence relating to each of these areas is considered in turn.

The labour market is one area where, in addition to a volume of work at the UK level, there is a substantial amount of Scotland-specific research. In recent years, several large quantitative studies have focused on specific aspects of the labour market in Scotland.

Migrants' education levels

The skills of migrants coming in to a country, relative to people born in that country, are important in determining the potential contribution they can make to the economy. Educational attainment is a useful proxy for skills (Saleheen and Shadforth, 2006:379), and the evidence indicates that, in general, migrants are more likely to have educational qualifications than the UK-born/Scottish-born population.

At the UK level, Wadsworth et al (2016) examined the impact of migration in the context of the referendum debate on the UK's membership of the EU. The study used Centre for Economic Performance analysis of the 2015 Labour Force Survey ( LFS). The researchers found that EU migrants are, on average, more highly educated than the UK-born population: about 43 per cent have some form of higher education, compared with only 23 per cent of the UK-born. Only 15 per cent of EU migrants left school at 16, compared with 44 per cent of the UK-born population (2016:4).

In Scotland, the 2014 study by Bell et al, which explored migration policy in relation to alternative constitutional outcomes for Scotland, involved secondary analysis of Annual Population Survey ( APS) data from 2009-2012 to examine labour market characteristics of the migrant population. 2012 APS data show that 13.1 per cent of the Scottish-born workforce in Scotland are graduates, and that this is a lower percentage than any migrant group. Examples given in the paper are 18.6 per cent (people born in Poland) and 49.1 per cent (Indian-born migrants) (2014:318). The authors found that the selectivity effect of migration policy has resulted in 'Scotland's immigrant population adding to the country's human talent by comparison with the Scots-born population' (2014:318).

Analysis of 2011 Census data by Scottish Government analysts indicates that almost half (48 per cent) of people who migrated to Scotland from EEA countries in the ten years prior to the Census have degree-level qualifications, as do 60 per cent of recent migrants from non- EEA countries. Approximately 40 per cent of EEA and non- EEA migrants who have lived in Scotland for more than ten years, and migrants born in the rest of the UK, are graduates. This compares with 22 per cent of people born in Scotland. Census data also indicate that approximately 10 per cent of recent EEA and non- EEA migrants have no qualifications, compared with about a quarter of established EEA and non- EEA migrants and 29 per cent of people born in Scotland (Scottish Government, 2016a:23).

Although the percentages vary, depending on the data sources used, and the ways in which migrant groups are disaggregated, it is clear that the majority of non- UK migrants arrive with qualifications and skills. However, the measurement of migrants' education levels is often complicated by the comparison of educational qualifications across countries (Nickell and Saleheen, 2008:3).

This difficulty is highlighted by Moskal in a qualitative study of new labour migration from Poland to Scotland (2013). The author found that, for people coming to Scotland with higher levels of educational and professional qualifications, the possibility of transferring formal qualifications may be a problem because 'this form of institutional cultural capital is often not recognised in the new society and thus devalued.' Moskal points out that many Polish migrants are working in jobs below their education levels, a situation they accept only because it is seen as temporary (2013:370). In a qualitative study of Polish migrants working in low-skill jobs in Scotland, Weishaar (2008) also found that respondents with post-school qualifications felt overqualified for the jobs they were doing, including 'cleaning, babysitting and catering.' Weishaar draws attention to the frustration migrants in the study felt because they were not able to work in the job for which they had spent many years studying (2008:1253).

The Scottish Parliament Equal Opportunities Committee 2010, charged to explore the impact and contribution of migrant populations within Scottish society, summarised the evidence considered in relation to the qualifications and skills of migrants:

'The Committee was told by a number of witnesses that many migrants who arrive in Scotland possess the relevant skills, experience and qualifications to enter more highly skilled jobs, but that qualifications obtained abroad are not recognised by employers or educational institutions. Some migrants therefore gave up pursuing a particular career and either take up employment in unskilled jobs or they may decide to leave Scotland altogether and take their skills and expertise with them' ( SPEOC, 2010: para 236).

Migrant employment

There is a good deal of recent, robust evidence relating to migrant employment. The key finding is that the majority of migrants to Scotland and the rest of the UK are in some form of employment. However, they may not be in employment that is a good fit for their skills and qualifications.

In relation to the contribution of migrants, the Scottish Parliament Equal Opportunities Committee (2010) heard evidence that:

'Many jobs would not be done without migrant workers. They keep public transport running, provide badly needed health care; and without their work, many farms and factories could not produce the goods they sell. Migrants in the health service and social care services make a critical contribution to Scotland. Many sectors could not function without us' Overseas Nurses and Careworkers Network ( SPEOC, 2010: para 120).

Scottish Government analysis of Census 2011 data found that 72 per cent of recent EEA migrants to Scotland were in employment (full- or part-time or self-employed), at the time of the Census, along with 56 per cent of established EEA migrants. 50 per cent of non- EEA recent and 60 per cent of non- EEA established migrants were in employment. This compares with 57 per cent of people born in Scotland and 60 per cent born in the rest of the UK (2016a:24).

The Bell et al (2014) study discussed earlier in this chapter included analysis of economic activity among four population groups: those born inside and outside the UK, those resident in Scotland and the rest of the UK. As Table 3.1 shows, more than 68 per cent of Scotland's non- UK migrant population were employed at the time data were collected (2009-12). However, in both Scotland and the rest of the UK, the proportion of the working age population in employment was higher among the UK-born than among those born outside the UK. The authors suggest that part of the explanation lies in the greater numbers of non- UK migrants who are students: they make the point that international students are a relatively more important source of migrants to Scotland than to the rest of the UK, as can be seen in the table.

Table 3.1: Economic activity among those of working age (per cent)

	Rest of the UK		Scotland
	UK-born	Foreign-born	UK-born	Foreign-born
Employed	70.9	66.2	71.0	68.1
Unemployed	6.0	6.7	6.2	5.9
Other 'inactive'	13.4	17.5	14.0	13.0
'Inactive' student	5.6	7.6	4.6	11.1
Retired	4.1	1.9	4.2	1.9

Source: Office for National Statistics: Annual Population Survey, 2009-2012. Reproduced from Bell et al (2014) Labour migration policy and constitutional change in Scotland

Bell and colleagues suggest that international students have economic significance in terms of their contribution to Scotland's exports, to future trade linkages, and to enhancing Scotland's productive potential, if they choose to stay in Scotland when they have finished their studies. This provides an argument for some differentiation policy (as was the case with the 'Fresh Talent' initiative which ran from 2005 to 2008 and offered overseas graduates from Scottish universities the opportunity to stay on for two years following graduation to seek employment) (2014:318). Bell et al argue that, since international student entry to the UK has been included on the UK Government-imposed cap on migrant numbers, UK policy is currently holding Scotland back ( ibid:319); see also Tindal, Findlay and Wright, 2014:11).

Several detailed assessments of the impact of migration on the labour market in the UK (and Scotland) have been conducted in recent years. Spreckelsen and Seeleib-Kaiser (2016) used LFS data from 2004-2014 to examine the UK labour market integration of recent young EU migrant citizens (aged 20-34). The authors found that young EU migrants appear well integrated in terms of employment, with higher employment rates than their UK peers. However, young migrants work, on average, longer hours than their UK peers and are less likely to work on permanent contracts, indicating that they are less well integrated in terms of job security and quality (2016:26).

Work by the Migration Advisory Committee ( MAC) in 2014 specifically examined the growth of EU and non- EU labour in low-skilled jobs, and its impact on the UK. The MAC analysis was based on a combination of desk-based research and evidence gathered from a range of sources, including the APS and the LFS. Findings indicate that, in 2013, there were around 13 million people in low-skilled jobs in the UK, of whom 2.1 million were foreign born, and one million of whom had come to the UK within the last ten years. Almost sixty per cent of migrants in low-skilled jobs in the UK had come from non- EU countries ^[13] (2014:6-7).

Although not mentioned in the MAC report, it should be noted that non- EU migrants are unable to obtain working visas in low-skilled occupations and, where migrants in this group are found to be in low-skilled work, it is because they have been granted indefinite leave to remain in the country, or have achieved citizenship, or have come to the UK on non-working visas. (See Annex 1 for a full definition of the tiers in the points-based visa system which operates under the UK's current migration regime.)

Another examination of migrants in low-skilled employment was carried out by Rienzo in 2015, using LFS data from 2014. The study found that, in 2002, there was only one low-skilled occupation (food preparation trades) in the list of top ten occupations with the highest share of non- UK born workers. In 2014, there were ' at least five low-skilled occupations on this list.' These included elementary process plant; cleaning and housekeeping; and assemblers and routine operatives (2015:4).

Rienzo's work does not differentiate between groups of migrants. However, Springford (2013) also used LFS data to examine the proportion of UK-born, Western Europeans and A8 nationals in different occupations. Chart 3.1 below (reproduced from Springford's report (2013:3)) shows that Western Europeans, tended to be working in more highly skilled jobs than the average person born in the UK. By contrast, a high proportion of A8 nationals were working in skilled and elementary trades, manufacturing, and elementary services jobs.

Chart 3.1: Occupations of people working in the UK

Source: Labour Force Survey, 2004-2012. Reproduced from Springford, J (2013) Is immigration a reason for Britain to leave the EU? Centre for European Reform, October 2013

In Scotland, Scottish Government analysis of Census 2011 data showed a similar pattern, with recent EEA migrants (the A8 group) most likely to work in elementary occupations (30 per cent of adults in employment) and established EEA migrants and people born in the UK more likely to be employed in jobs at the higher end of the occupational spectrum. Data are disaggregated slightly differently, so are not directly comparable, but the findings relating to recent EEA migrants is particularly striking (2016a:26).

Findings from two further studies of migrant employment outcomes in Scotland are similar to evidence from the UK as a whole in relation to the concentration of migrants at the top and bottom of the occupational scale. Vargas-Silva used LFS data up to 2012 to provide an overview of the characteristics and labour market outcomes of migrants in the Scottish labour market (2013). The author found that non- UK-born workers in Scotland tend to concentrate more in professional occupations (such as chemists, electrical engineers, pharmacists, solicitors) and elementary occupations (farm workers, cleaners, messengers and car park attendants, for example) relative to workers born in the UK (2013:5).

As Vargas-Silvas points out (2013) the sample of non- UK workers in the LFS who are resident in Scotland is small and, although the characteristics of all non- UK workers living in Scotland 'are inferred from the responses of those non- UK-born Scotland residents who are interviewed,' the estimates are still subject to 'substantial margins of error' (2013:2). However, because findings relating specifically to Scotland are similar to findings for the whole of the UK, it is likely that they are relatively robust.

The study by Bell et al (2014), also using LFS data up to 2012, found no substantial differences between the earnings of the UK-born and foreign-born residents in the rest of the UK. However, while the UK-born living in Scotland had a broadly similar income distribution to that in the rest of the UK, the foreign-born working in Scotland earned less than the other three groups ' from around the fourth to the eighth income decile.' The authors suggest this indicates that relatively few foreign-born workers are working in 'middle level' occupations in Scotland (2014:319-320).

In recent years, several qualitative studies have focused specifically on the labour market in remote and rural areas of Scotland, although these have generally addressed the experiences of migrants, employers and service providers rather than specific impacts of migration. De Lima (2012) found that Central and Eastern European workers are predominantly employed in hospitality, agriculture and food (2012:208). This study develops themes identified in an earlier study by De Lima and Wright (2009), who noted that the impact of migrant workers in rural areas has been primarily in low-skilled sectors, where the flexibility displayed by migrant workers is seen (presumably by the researchers) 'as an example of the labour market working well' (2009:393).

Research by Findlay, Geddes and McCollum (2010) examined differences in the use of migrant labour (particularly A8 migrant labour) following the onset of the recession. They used Worker Registration Scheme ( WRS) data (2004-09), ONS quarterly Employee Jobs Series, 2009, and primary research involving interviews with 60 employers of migrants. The authors found that A8 migrants tend to be concentrated in particular sectors of the Scottish labour market, with over half of all WRS registrations in hospitality and catering, agriculture, and food processing sectors (2010:306). Findlay et al report that 'A8 migrants make up a much larger proportion of the workforce in agriculture than they do in any other sector in Scotland' (2010:314/315).

A study by Vergunst (2009) focused on specific jobs within the agricultural sector in Scotland. The author points out that the majority of migrants are engaged in seasonal agricultural work, and that migrant labour 'is a prerequisite for financial viability of the 'other crop' sectors, encompassing soft fruits, vegetables for human consumption, and flowers and bulbs' (2009:261). However, in addition to seasonal jobs, some migrant workers undertake all-season agricultural work in the poultry and pig sector, for example (2009:261).

This section has focused on the work that migrants are employed to do. It is worth contextualising this by highlighting that migrants often fill labour market gaps created by a mismatch of skills, a mismatch between the types of jobs that unemployed people are willing to take on, and existing vacancies in the region, or/and information deficits which fail to match people with appropriate qualifications with relevant existing vacancies (Boswell et al, 2004).

Employers

This section of the chapter examines the evidence on migrant employment from the perspective of employers. It is an area where there is robust evidence, using a range of methodological approaches, relating specifically to the labour market in Scotland.

In their 2009 review, Rolfe and Metcalf reported on a body of Scotland-specific evidence to suggest that employers value migrants for the positive traits they bring to the workplace. These include punctuality, reliability, flexibility, productivity, work ethic, their willingness to accept low pay, poor conditions and fluctuating hours, their active approach to seeking work and positive integration with the existing local workforce (2009: para 3.8).

It is worth emphasising that while migrants' willingness to take lower wages and, potentially, accept poor working conditions, might be appreciated by employers, such willingness may not be good for migrant rights, or for the UK-born workforce, who may risk being undercut. The 2009 research by Vergunst, mentioned above, also focused on the preference for migrants expressed by employers in low-skilled sectors, while noting that migrants are likely to move into higher skilled, better paid jobs than seasonal agriculture, if and when these exist (2009:262). The Vergunst study emphasises the view of industry experts that, with improving economies in accession countries, the economic incentive to work in Scottish agriculture is likely to dwindle (2009:261).

The 2010 research by Findlay, Geddes and McCollum, discussed above, examined differences in the use of migrant labour following the onset of the recession. The researchers conclude that differences between agriculture and other sectors of the economy suggest that A8 migration serves two different functions in the Scottish labour market:

• Migrants act as a flexible supply of labour to be drawn on in times of economic boom, when labour demand exceeds local labour supply. In sectors where this is the main role for migrants (construction, for example) demand for migrant labour is likely to be cyclical and highly sensitive to economic cycles. Inflows of migrant labour would drop substantially during recession, when migrants are no longer required to boost the labour supply.
• In sectors such as agriculture, where employers find it difficult to source local labour, regardless of prevailing economic conditions, overseas workers form a more important part of the workforce. In the main, demand for this seasonal migrant labour remains steady during recession (2010:317).

More recently, research by McCollum et al (2015) used a literature review, interviews with a small number of employers and an online survey completed by more than 700 employers to examine the preferences and role of employers in the political economy. As in earlier studies, McCollum and colleagues found that migrants were seen as useful to fill labour shortages, especially in rural areas and where local labour was considered to be of poor quality. At the higher end of the labour market, migrants were described as being essential to address sector-specific skills shortages (such as healthcare, and oil and gas) and to act as a catalyst for growth for multinational companies with operations in Scotland (2015:10).

Qualitative research by the Department for Business Innovation and Skills ( BIS) in 2015 investigated the ways in which migrants contribute to the performance of the individual businesses they work for. The research interviewed 80 businesses across the UK and identified four key impacts:

• Skills - the skills held by migrant workers allowed businesses to expand their workforce, to fulfil existing contracts and take on more work. Employees with skills that go beyond the role in which they are employed bring additional benefits
• Innovation - migrants can facilitate innovation in business, particularly in sectors which generate creative products and in professional services where higher skilled employees frequently face new technical or creative challenges
• Knowledge sharing - this was found to impact directly, through the upskilling of co-workers and expansion of a business's knowledge base, and by facilitating other impacts such as training, innovation and international connections
• Migrants' connections - used to drive expansion in new and existing markets and improve client and customer relationships.

However, some businesses reported challenges relating to language and integrating migrants with UK-born co-workers in teams (2015:4-8).

The BIS research does not distinguish between migrant groups. However, Shubin and Dickey's 2013 study explored theoretical and practical issues in relation to the integration of Eastern European migrants in North East Scotland (Aberdeenshire). Aberdeenshire was selected for its prosperity, the fact that Aberdeen is 'widely acknowledged to be the European capital of the oil and gas industry' and because Aberdeenshire 'has experienced the highest rate of net in-migration from outside Scotland, as a proportion of the resident population, of any British county' (2013:2962), although the researchers' sources are not identified.

The research included in-depth interviews, public meetings and a survey of migrants which received over 200 responses, and a range of urban and rural locations. Shubin and Dickey found that the 'ready availability of a migrant workforce has transformed local employment landscapes' in North East Scotland. Migrants are treated by many Scottish employers as temporary workers to be hired as flexible labour through employment agencies, ' using a 'try-before-you-buy' approach' (2013:2973). The authors claim that their study endorses earlier findings that successive waves of migration have changed work organisation and employment relations in the host communities. As noted earlier, this is not necessarily a beneficial form of flexibility.

Research in 2014 by Tindal, McCollum and Bell explored the question 'to what extent does UK immigration policy currently meet the needs of Scottish employers?' Analysis drew on primary data from an online survey of employers (over 700 responses), and interviews with employers and industry representatives in key economic sectors in Scotland. Employers were 'emphatic' that Scotland's needs are not different from the rest of the UK, but that the rest of the UK (including Scotland) is different from London and South East England (2014:25).

The study by Tindal and colleagues includes a focus on the upper end of the skills spectrum. It highlights the observation by the director of a recruitment firm that the harmonisation of regulation across EU member states has allowed employers to access skilled labour which is in short supply in Scotland. The example given is that Romania has the same standards of dentistry as the UK. 'Now Romanian universities have taken the London University course and exams and it's totally replicated. It's a five year course and every person speaks English for that five years. So when they come out, academically they are absolutely 100 per cent fit for the role' (2014:21).

The research identifies three overlapping ways in which the freedom of EU citizens to live and work in the UK without restriction have benefited Scottish employers:

• Policies have allowed employers to recruit in great enough numbers, without restriction, from the EU, and expand their businesses
• Migrant labour is mobile and can be used to meet short-term labour demands
• Employers are legally able to recruit directly from the EU and move labour where it is needed (2014:26).

It also highlights ways in which UK immigration policies have restricted employers:

• Along with other Temporary Migration Schemes, the long-running Seasonal Agricultural Workers Scheme ensured a guaranteed labour force for agriculture; its termination (in 2013) means that employers need to rely on migrants wanting to come to the UK, and to work in specific sectors and in particular businesses
• There is uncertainty about whether companies will be able to recruit the same numbers in the future as in the past
• When applying quotas, the immigration system fails to recognise the specialist nature of particular jobs and mobility within organisations (2014:28-29).

Impact of migration on the wages and employment of UK-born workers

In addition to the skills and qualifications that migrants possess, the impact of migration on the labour market depends on the skills of existing workers, and the characteristics of the whole economy. As Ruhs and Vargas-Silva argue in a 2015 Migration Observatory briefing paper, it is important to distinguish between the effect of migration on the 'average wage of all workers in the economy, and on the wages of different groups of workers along the wage distribution' (2015a:2).

The 2012 MAC report on impacts of migration includes a useful summary of the empirical studies that have estimated the impact of migrants on UK average wages and the UK wage distribution (2012:66/67). The MAC report concludes that studies estimating the impact of migrants on UK wages have generally found little or no impact on average wages. Results from Dustmann et al (2005), Dustmann et al (2008), Lemos and Portes (2008), Nickell and Saleheen (2008), Reed and Latorre (2009) and Nathan (2011) suggest that an increase of 10,000 in the number of migrants in the UK changed average wages by between -£2 and +£2 per year (2012:59). In terms of wage distribution, the MAC report notes that in the majority of these studies, migrants were found to increase wages at the top of the UK wage distribution and reduce wages at the bottom (2012:59).

The available research indicates that any adverse wage effects of migration are likely to be greatest for resident workers who are themselves migrants. This is because the skills of new migrants are likely to be closer substitutes for the skills of migrants already employed in the UK than for those of UK-born workers (Ruhs and Vargas-Silva, 2015:4). Manacorda, Manning and Wadsworth (2012) conducted detailed analysis of data on wages, employment and education levels using the LFS and General Household Survey (mid-1970s to mid-2000s). Their aim was to gain a better understanding of why the significant rise in migration to the UK over that period had no appreciable effect on the average wages and employment of UK-born workers. The authors found that 'within narrowly defined age-education cells, immigration depresses the earnings of previous immigrants relative to the native-born, suggesting imperfect substitution between natives and immigrants in production' (2012:145). Thus, the main impact of increased migration in the UK appears to be on the wages of migrants who are already here.

As noted above, Spreckelsen and Seeleib-Kaiser (2016) used LFS data from 2004-2014 to investigate the labour market integration of recent young EU migrant citizens and the potential effects of the post-2008 economic downturn. They found that migrants from the EU accession countries have, on average, lower gross hourly wages than their UK peers (about 20 per cent less). However, Bulgarian and Romanian EU migrants have higher hourly wages than migrants from the other accession countries. The authors argue that this is possibly due to transition arrangements restricting the freedom of movement largely to high-skilled workers and the self-employed from Bulgaria and Romania until the end of 2013 (2016:18/19). The authors conclude that 'across our analyses there seems to be little change other than the compositional change, between the pre-/post-crisis labour market integration of youth migrants' ( ibid:27).

In a Centre for Economic Performance briefing on policy issues in the May 2015 General Election, Wadsworth investigated Annual Population Survey data between 2004 and 2012 to explore the impact of the recession. He did not find a robust correlation between changes in wages of the UK-born (either all or just the less skilled) and changes in 'local area immigrant share over this period' (2015:9).

The subsequent 2016 paper by Wadsworth et al highlights the body of research examining the effect of migration on jobs and wages in the UK, and the shared conclusion that the large increase in migration 'has not significantly harmed the job and wage prospects of UK-born workers.' However, the authors acknowledge that much of the work was conducted prior to the 'most severe economic downturn for 80 years.' Wadsworth and colleagues used analysis of the LFS up to 2015 to conclude that most people's wages fell during the recession and that the fall happened while EU migration was rising, 'but equally the big gains in real wages for UK workers were experienced at a time when EU immigration was also rising. So the cause of the fall of wages is the impact of the Great Recession - not immigration' (2016:6).

At the Scotland level, Pouliakas et al (2012) modelled the effects of migration on regional economic performance and wage distribution. This research includes the East Highlands of Scotland as one of three distinct remote regions of the EU. The authors argue that small regional economies lack flexibility and tend to be less diversified in their productive activities relative to a national economy. The Scottish region's exports, for example, are ' dominated by a particular manufacturing activity'. They found that an increase (decrease) in total labour supply is associated with a reduction (rise) in the region-wide wage of labour. So the substitution effect is stronger when there is less diversity. Specifically, a 10 per cent increase in labour decreases the wages of both skilled and unskilled workers, by 9-12 per cent in the Scottish region (2012:14).

There is relatively little evidence that migration has caused statistically significant displacement ^[14] of the UK-born population from the labour market in periods when the economy has been strong. However, a literature review carried out by economists from the Department for Business, Innovation and Skills and the Home Office (Devlin et al, 2014) found evidence of some labour market displacement in recent years when the economy has been in recession. Displacement effects were also more likely to be identified in periods when net migration volumes are high. In addition, where displacement effects were observed, these tended to be concentrated on the UK-born with low levels of skills. Evidence also suggests that where there has been a displacement effect from a particular cohort of migrants, this dissipates over time - that is, any displacement impacts from one set of new arrivals gradually decline as the labour market adjusts (2014:4).

As part of the analysis for the 2012 MAC report on the impacts of migration, researchers examined the association between migration and the UK-born employment rate, using data from the LFS between 1995 and 2010. The study found that a one-off increase of 100 in the inflow of working-age non- EU migrants was associated with a reduction in the employment of 23 UK-born workers over the period, although no statistically significant effects were found in relation to EU migrants. The authors point out that, although they controlled for differences in the UK-born employment rate across regions, which might affect the location choices of both UK-born people and migrants, results may still be influenced by a number of factors such as regional labour demand shocks and measurement error. Bearing in mind these caveats, they stress that their findings should be considered 'as estimating an association between migration and the native employment rate rather than the impact of migration on the native employment rate' (2012:62). They do not comment on the difference in findings relating to EU and non- EU migrants, which appear somewhat counterintuitive, given that non- EU migration is expected to be more selective based on skills and occupational shortages in the labour market.

The MAC analysis also suggests that the likelihood of a negative impact of migration on the employment of UK-born workers is greatest during economic downturn. In addition, findings suggest that it is only recent migrants (in the UK for less than five years) who are associated with the possible displacement of UK-born workers (2012:63).

Lucchino et al (2012) provide a useful summary of the literature on the impact of migration on 'native labour market outcomes.' They note that a popular strand in the literature compares employment changes across local areas with different migrant shares, to identify the impact of migration. The papers mentioned are all from 2005-2008, so are not investigated again in this review. However, the authors assess that there is no strong evidence of an overall effect of migration on aggregate employment and unemployment, and no significant effect of migration from A8 countries on either unemployment or wages (2012:3).

The authors also highlight that papers have tried to identify the impact of migration by splitting the population into segments based on workers' characteristics (such as age, gender and skill levels) in order to estimate the impact of migration by comparing changes in employment across segments of the population with different migrant inflows. However, they argue that this methodology rests on the assumption that migrant and native workers in each segment are substitutes. 'This is a central consideration, as the impact of immigration on the labour market outcomes of a given segment of the native population is likely to be more adverse (beneficial) the greater the substitutability (complementarity) between migrants and workers in that same segment' ( ibid:3).

Lucchino et al contributed to the existing body of evidence by using data on National Insurance Number ( NINo) allocations to adult overseas nationals entering the UK as a measure of migrant inflows. The authors claimed theirs was the first empirical paper to make use of these data to examine the employment effects of migration and, by using NINo registration data up to 2010/11, they hoped to shed light on the impacts of the recession. Findings confirm the general lack of an aggregate impact of migration on unemployment, supporting previous findings in the literature. In addition, by using more recent data, Lucchino et al found no evidence for a more adverse impact of migration during the recession. However, there are difficulties associated with conducting analysis on migrant economic activity using NINo data, as illustrated by a study by Saggar et al (2012). The authors investigated the time between arrival and registration for migrants registering in 2010/11 by world region of origin and found that 8 per cent of EEA and 18 per cent of non- EEA migrants registered more than two years after they arrived in the UK (2012:42).

In a study of the economic and demographic consequences of large-scale migration for the UK, Rowthorn (2014) makes the point that between the first quarter of 2008 and the first quarter of 2010, the number of UK-born people in employment fell by 700,000, or nearly 3 per cent. During the same period, the number of 'foreign-born workers' in employment remained virtually constant. Taking a longer view, he notes that, despite 'strong overall employment growth,' the number of UK-born people with a job in early 2014 was still below its 2005 peak, while the number of non- UK born workers in employment had increased by around 1.5 million. Rowthorn concludes that, although such aggregate comparisons do not prove definitively that migrants have displaced UK-born workers, 'they do suggest there is a case to answer' (2014:23-24). However, Rowthorn's work does not differentiate between migrant populations or length of time in the UK.

Ruhs and Vargas-Silva provide a summary of relevant evidence at the UK level in their 2015 Migration Observatory paper. Findings from their research concur with those of other authors, in that they did not find a significant impact of overall migration on unemployment in the UK, although evidence suggests that migration from outside the EU could have a negative impact on employment of UK-born workers, especially during an economic downturn. The authors conclude that for both wages and employment, short-run effects of migration differ from long run effects: 'any declines in the wages and employment of UK-born workers in the short run can be offset by rising wages and employment in the long run' (2015:2).

Impact of migration on public finances

Key points

• There are variations in estimates of the fiscal contributions made by migrants. Although the evidence is contested, research indicates that those from EEA countries, particularly those who came to the UK after 2000, have made a more positive fiscal contribution than non- EEA migrants, and people born in the UK.
• Migrants from the EEA are less likely to claim out-of-work benefits, but more likely to claim in-work benefits, compared with people born in the UK. There is likely to be a connection between the level of in-work benefit claims and the types of low-paid jobs taken by migrants.
• Work in the UK and the 27 OECD countries has estimated that, where migrants have had a fiscal impact, it has rarely exceeded plus or minus 0.5 per cent of GDP.

Measuring the impact of migration on the public purse is a complex task. The net fiscal impact is typically estimated as the difference between the taxes and other monetary contributions made by migrants to public finances and the costs of the public benefits and services they receive. However, this impact depends on the characteristics of migrants, their impacts on the labour market, their engagement with (and the characteristics and rules of) the social security system, among other factors. In addition, existing estimates of fiscal impact are limited due to a lack of accurate data on non-monetary contributions (such as the provision of 'free' caring and support services by family members). A significant number of assumptions must be made, and results tend to change based on these assumptions (Vargas-Silva, 2015a:2-3). In addition, as noted by Vargas-Silva (2015) migrants deliver public services as well as consuming them. The author argues that 'it may be possible to deliver services in the public sector at a lower cost because of the availability of migrants willing to work at a lower wage' (2015:8).

A number of relevant papers have been published in recent years, although the evidence is all at the UK level. It is difficult to draw consistent messages from findings from individual studies, however, as these have often been contested. This section provides a brief summary of the available evidence. While it is not possible to conclude whether the fiscal impact of migration is generally positive or negative, there are indications that recent EEA migrants are the most likely to have a positive fiscal impact, and that all non- UK migrants are likely to make a more positive contribution than people born in the UK.

Migrant contributions to public finances

There are two main approaches to conducting analysis of migrant contributions to public finances: static - based on a single year and using historical data; and dynamic - providing a forward looking perspective over the lifetime of migrants. In the UK to date, all significant studies have used a static approach (Vargas-Silva, 2015a:3-4). However, one recent study claimed a dynamic approach to its analysis of historical data (1995-2011). Findings indicate that migrant contributions are positive, particularly in relation to EEA migrants who have entered the fiscal system since 2000 (Dustmann and Frattini, 2013).

The assumptions made by the authors in their calculations (in relation to the allocation of revenue streams, for example) were challenged by two subsequent studies. The (anonymous) authors of a paper by Migration Watch (2014) examined the claims made by Dustmann and Frattini, and suggested that they had exaggerated the revenues the Government obtains from migrants and underestimated the cost of benefits and tax credits provision to migrants (2014:49).

Because the author(s) of the Migration Watch study are not named, and because Migration Watch is not itself a research organisation, it is difficult to gauge the robustness of the analysis. However, Rowthorn (2014) also assessed that the estimates of migrant-generated revenue by Dustmann and Frattini were too high. He argued for the inclusion of a British worker displacement adjustment, on the basis that migration had at least a temporary impact on the level of 'native employment' during the recession of 2007-11 (2014:53).

Dustmann and Frattini responded with an update to their analysis (2014). They also provided a further breakdown between A10 countries and the rest of the EEA. The authors estimated that recent EEA migrants (A10 and other EEA) made a net fiscal contribution of approximately £20 billion between 2001 and 2011. This compared with a negative fiscal contribution of approximately £616.5 billion by people who were UK-born, over the same period.

Dustmann and Frattini also calculated that, even throughout the years of greatest economic instability (2007-2011) recent A10 migrants made a net contribution of almost £2 billion to UK public finances, while recent migrants from other EEA countries made a positive contribution of £8.6 billion (2014:620/21).

However, as Table 3.2 indicates, there is still a major variation in estimates of fiscal contributions that have been made in recent years.

Table 3.2: Estimates of the fiscal effects of immigration for the fiscal years 1995-2011 and 2001-2011 (billion, 2011 GBP equivalent)

	Net fiscal impacts
	UK-born	All migrants		Recent migrants
		EEA	Non- EEA	EEA	Non- EEA
Dustmann and Frattini (2013)
1995-2011	-604.5	+8.8	-104.1
2001-2011	-624.1	+9.0	-86.8	+22.1	+2.9
Dustmann and Frattini (2014)
1995-2011	-591.5	+4.4	-118.0
2001-2011	-616.5				+5.2
2001-2011 (A10)				+4.9
2001-2011 (rest of EEA)				+15.3
Migration Watch UK (2014)
1995-2001	-565.3	-13.6	-134.9
2001-2011	-586.2	-13.4	-116.8	-0.25	-27.17
Rowthorn (2014)
2001-2011	(Not shown)			-0.3	-29.7

Source: Vargas-Silva, C (2015a) The fiscal impact of immigration in the UK, The Migration Observatory at the University of Oxford, 27 March 2015, page 6, Table 2. ' UK-born' column from the original publications

Finally, HM Revenue and Customs ( HMRC) examined statistics relating to recently arrived non- UK EEA nationals subject to income tax and national insurance contributions, or receiving HMRC administered benefits. Based on the 2013-14 tax year, the 2016 work estimated that migrants in this group were subject to £3.11 billion income tax and national insurance contributions and received £0.56 billion in tax credits or child benefit, leading to a net fiscal contribution of £2.54 billion (2016:6).

Social Security ^[15]

One of the negative perceptions of migrants is that they are more heavily dependent on social security payments than people born in the host country. However, while there is a good deal of evidence relating to this issue, it is difficult to draw any consistent messages from the findings from individual studies. Drinkwater and Robinson (2013) summarise the international literature and highlight mixed results across a range of countries. The authors used LFS data (2004-09) to focus on the types of benefits that migrants to the UK tend to claim, as well as examining differences by area of origin (disaggregated by three European and four non-European categories). The research focused on country of birth as the defining variable although, as the authors pointed out, this necessarily includes both migrants who are recent arrivals and those who settled more than 50 years ago (substantial levels of migration from the West Indies during the 1960s, for example).

The research concludes that social security claims vary considerably by migrant group as well as by the type of benefit claimed in the UK. The authors found that Australasian, American and A8 migrants were the least likely to claim benefits, a finding they suggested could, to some extent, be explained by the characteristics of individuals from these groups, especially those from accession countries, who are more likely than people born in the UK to be in employment (2013:109). Drinkwater and Robinson also found that ' non-white migrants' were ' disproportionately more likely to claim welfare benefits.' They suggest that there may be 'cultural explanations' for this, but that 'an element of discrimination cannot be ruled out given the consistent findings across all groups' ( ibid:109). However, the paper does not make it clear what the authors mean by this.

Drinkwater and Robinson also argue that the relationship between migrants and welfare participation is not static: 'welfare payments will certainly have been heavily influenced by the recent recession and the subsequent impact on government finances. However, it is unclear what effect the recession will have had on the relative levels of welfare participation by immigrants in the coming years' ( ibid:109). In addition, the authors point out that their research was carried out too soon to detect the impact of the points-based system of migration, introduced in the UK in 2008.

A body of work by Dustmann and colleagues has investigated the topic of welfare receipt by migrants as part of their broader research on the fiscal impacts of migration. Dustmann, Frattini and Halls (2010) found that A8 migrants who arrived after EU enlargement in 2004 had a 59 per cent lower probability of being in receipt of welfare assistance than people born in the UK (decreasing to 49 per cent for those in the UK for longer than two years). The authors conclude that, even if A8 migrants were identical to UK-born people in their educational background, age and gender composition and number of dependent children, they would still be 13 per cent less likely to receive benefits (2010:15).

A later paper (Dustmann and Frattini, 2013) included slightly different figures, without referencing the earlier calculations: the researchers argued that recent migrants were 45 per cent less likely to receive state benefits or tax credits than people born in the UK. The authors acknowledge that this may be partly explained by migrants' ' more favourable age-gender composition.' However, even when compared to UK-born people with the same demographic characteristics, recent migrants were 21 per cent less likely to receive benefits. The authors also found differences between EEA and non- EEA migrants, with recent EEA being more than 50 per cent less likely than people born in the UK to receive state benefits or tax credits, compared with a 43 per cent lower likelihood for non- EEA migrants (2013:27-28).

This claim that recent EEA migrants are only half as likely to claim benefits or tax credits was challenged by the Migration Watch study which examined the findings of Dustmann and Frattini (2014). The study argues that in the context of establishing the fiscal cost, what matters is the amount people receive, rather than the number of claims made, and different benefits pay different amounts to different people. 'Recent EEA migrants are much more likely to receive tax credits than the UK-born population, and more likely to receive housing benefit, 'and these are likely to be paid at higher rates in view of their lower incomes' (2014:49).

More recently, a Migration Observatory report focused on the main policy issues and statistics on EU migrants' use of social security benefits in the context of the EU referendum of June 2016 (Sumption and Altorjai, 2016). The study analysed LFS and HMRC data to conclude that EU migrants are less likely to claim out-of-work benefits than people who are born in the UK. The authors argue that this is driven primarily by 'lower use of incapacity benefits (1.7 per cent of claims in February 2015) but also to some extent by lower use of job seeker benefits (4.8 per cent of claims)' (2016:6).

Sumption and Altorjai found that the picture for in-work benefits, such as tax credits and housing benefit, is different . Noting that actual figures vary, depending on the source of data, the authors calculate that it is reasonable to assume that the share of recent arrivals from EU countries receiving tax credits in early 2014 was 'between 10 per cent and 20 per cent' ( ibid:9). (The share of UK-born people receiving tax credits was 10 per cent.)

The authors acknowledge that administrative data on housing benefit recipients 'are extremely limited.' However, they estimated that the number of EEA-born people who reported receiving housing benefit in 2015 was lower than the number who reported receiving tax credits (198,000 and 305,000 respectively) ( ibid:9).

The research commented on a statement made by the UK government in November 2015:

'… [in March 2013] between 37 per cent and 45 per cent of the EEA nationals (excluding students) who were resident in the UK having arrived in the preceding four years were in households claiming either an in-work or out-of-work benefit or tax credit' (Department of Work and Pensions, 2015:7).

The authors highlight that this estimate is higher than other available statistics. Their analysis of LFS statistics found that approximately 22 per cent of recently arrived EEA adults reported receiving a form of state benefit in early 2013. By 2015 this had fallen to 18 per cent. In other words, the DWP analysis suggested a rate of benefit receipt for recent EEA migrants 'about double the share found in the LFS' ( ibid:10).

The Sumption and Altorjai study also found that more than half of EEA-born adults who reported receiving tax credits in 2015 were working full time, with median earnings of £15,600; the largest numbers were employed in wholesale/retail trade, manufacturing and hospitality ( ibid:11/12).

Sumption and Altorjai drew attention to the most recent estimate from the UK government ( DWP and HMRC analysis of 2013-14 data): that £2.5 billion or 10 per cent of in-work tax credit and housing benefit expenditure was on claims involving EEA nationals. Out-of-work expenditure on claims led by EEA nationals made up £886 million, or 3 per cent of the total DWP out-of-work benefits bill ( ibid:15).

Although there are no Scotland-specific calculations of levels of benefits, a few qualitative studies have highlighted particular perspectives associated with social security benefit receipt. McGhee et al (2013) considered the implications for Polish migrants who move into areas already experiencing multiple social problems. In an estate where many people were described by Polish study participants as ' life-long welfare dependents,' the research found that Poles were often viewed in a positive light by resident Scots for their contribution to the labour market and associated taxes paid from wage labour. Consequently, Poles were not seen as competitors for jobs or houses, but were viewed as workers and payers of taxes, while their Scottish neighbours on the estate 'get the benefits' (2013:335).

One of the findings from Shubin and Dickey's qualitative work in relation to Eastern European migrants in Aberdeenshire (2013) was that many Eastern European migrants reject the idea of any state support in Scotland, because they see it as the sign of their lack of success. 'For many Eastern Europeans state support in their home countries was very limited, and using government help was linked to marginalisation. Consequently, migrants often reject any routes to success that are state sponsored - including job search in job centres and employment support by local government' (2013:2973).

Migrants' contribution to GDP

Work by the OECD indicates that migrants have had a broadly neutral impact in the 27 OECD countries over the past 50 years. In other words, the cost of whatever state benefits they received was largely covered by the taxes they paid. Where migrants did have a fiscal impact, it rarely exceeded plus or minus 0.5 per cent of GDP ( OECD Policy Brief, May 2014). Using data from 2007-09, published by OECD, Vargas-Silva reported that the fiscal impact of migration in the UK over that period (+0.46 per cent of GDP) was more positive than the fiscal impact of migration in 16 other OECD countries (2015a:8).

In their work on fiscal impacts, neither Dustmann and Frattini nor Migration Watch included a calculation of the implications of the cost/contributions of migration in terms of GDP. However, commenting on findings from the 2013 Dustmann and Frattini study, Rowthorn estimated that, although the total net contribution assessed by the authors between 2001 and 2011 are 'large numbers in absolute terms', the net fiscal contribution is in the range of -0.7 to +0.3 per cent of GDP, depending on how it is measured. Rowthorn advised that the Dustmann and Frattini finding that post-2000 migrants from the EEA have generated a large fiscal surplus should, therefore, be seen in perspective (2015:68).