2. An overview of the evidence base
Scotland-specific evidence on the impacts of migrants and migration is somewhat limited. Research tends to be descriptive and/or theoretical and, as noted in Chapter 1, to focus on the needs and experiences of migrants, or particular populations of migrants, rather than their impacts on Scotland as a receiving country. The evidence base has, however, been strengthened in recent years, and a number of studies are relevant to this review.
While there are relatively few academics publishing in the field of migration impacts in Scotland, several experts have published widely. In addition, organisations such as the Economic and Social Research Council-funded Centre for Population Change have produced a number of useful studies at the Scotland level.
Work by the Scottish Parliament Equal Opportunities Committee in 2010 (Inquiry into migration and trafficking) explored the impact and contribution of migrant populations to Scottish society. The Committee's report noted that, while migration is a reserved issue, the ' Scottish Government must have a role in formulating and developing migration policy given the impact on devolved responsibilities such as healthcare, education and the justice system' (2010: para 685). A range of experts, including academics, gave evidence to the Committee. Throughout this review, relevant opinions and experiences offered to the Committee are included in boxes to supplement the text.
The evidence base relating to economic and labour market impacts is more developed than evidence on public services, social and cultural impacts. In particular, there has been a good deal of work in recent years on the fiscal impacts of migration. At the UK level, researchers at the University of Oxford's Migration Observatory have conducted a range of relevant studies, and identified problems in the evidence base as part of that work. This review will not examine these studies in detail, but will locate the Scottish evidence base within the UK evidence base where possible. Several Migration Observatory briefing papers focus specifically on Scotland.
Information on migrants' contribution to the provision of public services in Scotland is still scarce. However, individual studies are beginning to fill the gaps in the evidence base, both at the UK and Scotland level.
The most comprehensive body of work to be conducted on the impacts of migration since 2008 is a study by the Migration Advisory Committee ( MAC), published in 2012. This included six specially commissioned research projects, workshops with academics, government officials and public sector bodies, and a stakeholder panel. Several of these projects make specific reference to Scotland. The focus of the study (and most of the contributing research projects) is on non- EEA migrants and, in particular, skilled workers and students. However, it is relevant to both economic and public service impacts, so will be referred to throughout this review. The focus on non- EEA migrants is also important in the context of the increased numbers of migrants to Scotland from India, Pakistan and China between the 2001 and 2011 Censuses (Table 1.2).
The 2009 review of the literature by Rolfe and Metcalf is referenced by many of the more recent Scotland-specific studies, so it is important to consider the evidence base in relation to the previous study commissioned by the Scottish Government. Although the scope, scale and focus of the earlier report was slightly different from the present review, the work of Rolfe and Metcalf will be referenced where it is helpful to do so.
There is a link between the scope and scale of research and the choice of research method. Large studies, which explore the extent of impacts across populations, tend to favour quantitative methods, while smaller studies use qualitative methods, or a combination of both. A number of studies include useful literature reviews.
Valuable insights on impact may be drawn from small research projects which explore the depth of experience in key populations or localities. Such studies involve migrants whose characteristics, experiences and motivations for migration are not necessarily shared by others, and may be specific to time and place. A number of informative qualitative studies have been conducted at a local level, with Glasgow and rural Scotland being particularly well represented.
About data sources
There is an identified and ongoing need for better data relating to migrants, both at the Scotland and UK level. However, as acknowledged by the National Records of Scotland ( NRS), migration is the most difficult component of population change to estimate. There is no comprehensive system which registers migration in the UK, including migration to or from overseas, migration to or from other parts of the UK or migration within Scotland. Therefore, estimates of migration have to be based on survey data and the best administrative data that exist. The NRS website includes a section on the availability and sources of migration data, and a spreadsheet detailing the strengths and weaknesses of different sources  .
The issue has been examined several times, most recently in 2013 by the UK Government Public Administration Committee  . (See also the UK Government's response  and the response from the UK Statistics Authority  .) Work continues to improve the availability and sources of data, and ONS are taking forward work to link HM Revenue and Customs and Department of Work and Pensions data to look at the characteristics of migrants.
The review did not set out to examine the data sources used in the studies in the evidence base. However, data source deficiencies are highlighted by researchers in a number of areas and, while these have not been explored further as part of this review, it is worth noting that they exist. For example, migration inflows to the UK from EEA and non- EEA countries may be examined using National Insurance Number ( NINo) registration datasets maintained by the Department of Work and Pensions. However, NINo data have shortcomings: they do not record the date a migrant arrives, but only the date they register onto the NINo system, in order to work or claim benefits. Research has indicated that more than a quarter of NINo registrations in 2010-11 took place more than two years after migrants arrived in the UK (Saggar et al, 2012).
The Annual Population Survey ( APS) is used by many of the studies to gain a better understanding of migrant experiences in the host country. The APS is a continuous household survey, covering the UK, with the aim of providing estimates between censuses of key social and labour market variables at a local area level. The APS is not a stand-alone survey, but uses data combined from two waves from the main Labour Force Survey ( LFS) with data collected on a local sample boost. Apart from employment and unemployment, the topics covered in the survey include housing, ethnicity, religion, health and education  .
The LFS is used by many of the studies that have aimed to gain a better understanding of migrant experiences in the labour market. However, in addition to issues relating to small sample sizes for migrants, the survey excludes short-term migrants and those who do not live in households. It also excludes halls of residence, thus missing many international students who may be legally working in the UK. Consequently, there are limitations to the way the data can be interpreted (Vargas-Silva, 2013:6).
About migrant groups
The studies included in this review recognise that migrants are not and should not be considered as a homogenous group, although it is not always possible to distinguish between migrant groups when findings are reported. Most recent research in Scotland relates to A8, primarily Polish, migrants. Some studies focus on all migrants, generally meaning 'all non- UK' migrants. Individual research projects also focus on specific communities (such as the Roma population in Govanhill).
The review excludes evidence relating to refugees and asylum seekers. Where studies include groups of refugees and asylum seekers in addition to other types of migrant groups, only data which relate to the in-scope groups are reported, at least as far as it is possible to disaggregate findings.