As researchers in the field of international migration, we confront the confusion that surrounds the subject each day. Concepts are unclear. Statistics are often inadequate. Reductionism is rife. Migration becomes "immigration"; immigration and asylum are conflated. There is a foghorn dialogue between the "pro" and "anti" lobbies.
The answer to the question "What is migration?" is not straightforward. Migration is a sub-category of the wider concept "movement," embracing various types and forms of human mobility from commuting to permanent emigration. We want to throw some light on the scale and characteristics of international migration as it affects the UK. We do not argue any view; we merely set out what is going on, as far as the data allow us.
Migration is the main component of population change in 26 out of 45 European states, including the UK. Economic globalisation, transnationalism and established diaspora communities have increased the circulation of people. New migrations have emerged, many of them short-term, taking advantage of opportunities to exploit niche labour markets. Shortages of skill and manpower at the top and bottom ends of the labour market have increased openness to foreigners. Unrecorded and illegal migrations continue to pose challenges, although there is no hard evidence that their scale is increasing.
Migrants themselves and a range of institutions—including government, employers, trade unions and human rights groups—have their reasons for promoting or trying to limit movement. Migration may be seen as an industry in which these various actors seek to maximise their objectives through a complex series of accommodations and compromises.
The migrations that do occur are the product of different processes and circumstances. For example, the right of a migrant to family reunion is often accepted as a basic human right and this facilitates the flow of spouses and children. Overseas student recruitment is usually left to educational institutions which are bound by no limits on numbers and wish to attract as many fee-payers as possible. In contrast, the movement of seasonal or low-skilled workers is often managed by state quotas, such as those in the agriculture and hospitality industries. Highly skilled migrants come by other routes such as corporate assignments. Different circumstances surround retirement migration, where moves are usually personal decisions, tempered by social security and pension rules.
Simultaneously, thousands of people each year return to their country of origin or migrate elsewhere. In 2003, for example, 106,000 Britons returned to live in this country while 191,000 emigrated. And it is not often commented upon that there are at least 3.5m Britons living in other OECD countries.
How is migration measured?
For most of the 19th and 20th centuries, more people left than entered Britain. It was only during the 1980s that an annual net inflow became the norm (see chart below), although in the early 1990s the country slipped into a small migration deficit again. The second half of the 1990s saw a sharp rise in inflows which has continued into the new millennium. We are currently in the period of highest ever recorded net inflow. More immigrants are arriving in the country than ever before. They include workers, students, family members, asylum seekers and others.
Although data provision has improved, it remains far from ideal. Particularly difficult to capture are short-term movements and changes of immigration status, as well as, most obviously, irregular or illegal migrations.
There are two main types of data. Survey data, like the International Passenger Survey (IPS) and Labour Force Survey (LFS) are based on small samples prone to error, are not comprehensive and are impossible to break down with any accuracy. Administrative data, on the other hand, such as numbers granted settlement or citizenship, are collected by government for purposes other than counting migrants. Both types are particularly weak at providing information about emigration. Yet we know from the treasury Pink Book, which analyses balance of payments data, that the economy gains more from remittances by its citizens working abroad than it loses by foreign workers sending money back home. In 2003, the net gain was around £1.5bn.
Periodically, governments evaluate their data. In 2003, the Office for National Statistics (ONS) completed a review of UK international migration data. The review concluded that there was little chance of major improvements in accuracy without heavy extra spending.
Electronic entry and embarkation controls have been proposed as a way of recording movements more accurately, using machine-readable passports or ID cards containing biometric data. There are various problems with such systems, including expense, the sheer practicalities given the 90m passenger arrivals in the UK in 2003 (and a similar number of departures), and freedom of movement within the European Economic Area (the EU states plus Norway, Iceland, Switzerland and Liechtenstein). At present only Australia has such a system for all entrants.
The headline figure
The media most often quote the figure of 150,000 immigrants a year—the net inflow for the last couple of years for which we have data, 2002-03. Because of the significance of this number it is important to understand how it is derived (see chart, right). It comes from the only source of both immigration and emigration data, the IPS, an annual survey of the roughly 180m passengers arriving at and departing from UK air and sea ports. Most people surveyed are visitors, but a small sub-category of migrants is identified. Immigrants and emigrants are defined as those intending to stay in the UK or leave it for a year or more, having lived outside the UK (for immigrants) or inside the UK (for emigrants) for a year or more. The IPS does not cover routes between the Irish Republic and the UK, but separate estimates are made of these flows.
In 2002, 2,477 migrants were surveyed on their way into the country and 768 questioned on their way out. These numbers were then extrapolated to produce a total flow (in and out) of around 700,000, rising to 872,000 after adjustments (see below) with an ultimate net migration gain of 153,000. The asymmetrical nature of the sample means that our knowledge about the scale and pattern of emigration is derived from interviews with fewer than 800 people. For additional information we must go to the immigration statistics of other countries. Although the ONS has gained a lot of experience over the years in deriving a gross flow that is as accurate as possible, annual fluctuations that are the product of sampling error—the standard error is about 4 per cent—rather than actual changes do occur.
Two immediate difficulties with the IPS data are that it uses a 12 month cut-off, so that anyone coming or going for a shorter period is not counted as a migrant, and it is based on people's stated intentions. This means that it is likely that they exclude people seeking asylum after entering the country, and those admitted as short-term visitors who are subsequently granted an extension of stay for a year or longer, for example as students or after marriage. We refer to these people as "switchers" and an adjustment is made to the statistics to take them into account.
The peak adjusted net inflow of migrants was recorded as 171,800 in 2001, falling to 151,000 in 2003, the latest year for which data are available. This headline figure is the difference between inflows and outflows, each of which may change from year to year. Hence in several years recently, net inflow has risen because of a fall in numbers of emigrants as well as an increase in immigration. Further, it reflects the migration of British citizens (in and out) as well as numerous foreign streams. Finally, it is affected by the balance between the IPS survey figure and the size of the adjustment made, neither of which is constant from one year to the next.
Let us look at how the net migration figure of 151,000 for 2003 was derived. First, an adjustment of 33,000 was made to take account of those whose intentions about length of stay changed. Second, there was a net loss of 85,000 Britons, the result of 106,000 coming in but 191,000 emigrating. Third, there was a gain of 236,000 foreign citizens, the difference between 407,000 entering and 171,000 leaving. The individual streams of foreign citizens may change from year to year: for example, the rise in net immigration in 1997-98 was principally because of a smaller net loss of Britons (more returning, fewer leaving) and by rising gains of EU and old Commonwealth nationals.
Foreigners and citizens
This statistical soup illustrates the complexity underlying one headline figure. Before returning to these flows we will first look at the UK in a European context and at the link between migration and citizenship.
In spring 2004, before EU enlargement, there were 2.857m foreign nationals living in the UK, 4.9 per cent of the total population. Forty-three per cent of them were European, of which 79 per cent were from the European Economic Area, within which citizens have the right to move freely between countries and take up work. Around a quarter were Asian, 17 per cent African and 10 per cent from the US. The largest group from a single country was the Irish (13 per cent), but their numbers have dropped sharply over the last decade. There were more women (53 per cent) than men.
Currently, the UK has about 12 per cent of western Europe's foreign residents, well behind Germany's 31 per cent (though this reflects the low rate of naturalisation there) and between France's 14 per cent and Italy's 9 per cent. The proportionate size of our foreign population is modest compared with some other countries, including Austria, Germany, Belgium and Ireland. In terms of the scale of inflows, the UK is in third place, after Germany and Spain, but is second to Germany in numbers emigrating. Most of western Europe is still recording net annual migration gains, with the UK occupying the number two spot after Italy. Across Europe, in 2003, after several years of increase, the trend in numbers of asylum seekers went into reverse, with 13 out of 18 countries showing declines, the UK having the largest percentage fall.
What is interesting in European migration today is that echoes of the guest-worker period of the 1960s can be heard. Temporary migration into sectors such as agriculture, construction and hospitality is a familiar theme. The lessons of history suggest that now, as then, temporary will become permanent. Moreover, this time eastern as well as western Europe is affected.
Of the UK's 2.857m foreign nationals, 1.445m were working, representing 5.2 per cent of all people in employment. Foreign nationals were more likely than Britons to be in the most skilled category (professionals and managers) and slightly more likely to be in the least skilled occupations. EU and especially North American citizens were particularly likely to be more skilled. The regional distribution of foreign workers was very uneven, two thirds of them residing in London and the southeast. Given the stability of this distribution for many years, it is likely that any increased flows of immigrant workers will gravitate mainly to this corner of the country, adding to regional growth but also congestion.
A measure of immigration sometimes used is the number granted settlement—the right to settle permanently (see right). During the 1980s and early 1990s, numbers fluctuated around 50,000 a year, since when they have risen from 58,725 in 1997 to a peak of 143,845 in 2003. Most of these were family members, not all living abroad; others were allowed settlement after a period of employment in the UK, after recognition as refugees or under exceptional leave to remain arrangements. About half in 2003 were from Africa (31 per cent) or the Indian subcontinent (21 per cent).
Naturalisation is often regarded as a measure of integration, the transition from being a foreign migrant to becoming a British citizen. (People granted right of settlement who do not become citizens are known as denizens.) Numbers granted citizenship have been rising, and reached 124,300 in 2003. Residence in the UK for a period of five years was the most frequent basis for grant of citizenship, 44 per cent; about a third were on the basis of marriage and a quarter to children. About a third were formerly citizens of an African country and a quarter were from the Indian subcontinent.
Economic migration: how big is it?
Political and media discussion makes much of "economic migrants," often as the Tweedledum to asylum seekers' Tweedledee. Economic migration is a very diverse category, covering everyone from au pairs to agriculture, construction and manufacturing workers, hotel, catering and cleaning staff, corporate managers, health professionals, footballers, ministers of religion, chick sexers and many more.
There are numerous officially recognised routes of entry for economic migrants. Each of these is managed to a greater or lesser extent and responds to different rules and conditions. Legal labour immigration is substantial—the total figure in 2003 was 238,600, the top five categories being: work permits and "first permissions" (85,300), working holidaymakers (46,500), EU citizens (35,000), seasonal agricultural workers' scheme (23,300), and au pairs (15,300). We have little idea how long these immigrants stay. Some of the schemes listed have time limits, such as the seasonal agricultural workers' scheme. Others, like the work permits scheme, are more elastic. EU citizens can stay indefinitely. (Visitors from about 100 countries to Britain still require a visa. About 1.6m people arrived in 2003 on visas, most of them for short stays. There is no longer any mechanism to check whether visa visitors have left the country.)
What the total figure for legal labour immigration will be in 2004 is not yet known. Between May and the end of the year, 123,000 citizens of the eastern European accession countries registered for work under the worker registration scheme, some 45,000 of whom were already here. Poles were the largest group (56 per cent), hospitality and catering (28 per cent) the main sector of employment.
The main mechanism for managing the entry of foreign labour is the work permit system. In recent years this has expanded considerably. The current government promotes economic migration to meet skill shortages and to increase the competitiveness of the economy. The work permit system is an instrument to enable this.
A work permit is granted to the employer on behalf of a named worker, to fill a vacancy that, in theory, cannot be filled by an EU citizen. The number of work permits issued has risen dramatically, from just under 33,000 in 1995 to 139,000 in 2004, the largest number ever recorded. But the figure needs interpreting. There are four main elements to the system: work permits, first permissions, extensions and changes of employment. Only the first of these includes foreigners living outside the UK. First permissions are permits issued for the first time on behalf of foreigners already in the UK, perhaps foreign graduates obtaining a first job with a British employer. Extensions allow foreign employees who already have work permits to extend their stay, while change of employment allows someone with a work permit with one employer to move to another.
Work permits and first permissions together measure how many new foreign workers have been employed. They were just under two thirds of the total in 2004. The rest were already working in the UK. Furthermore, work permits for foreign workers who are living abroad at the time of approval now account for only 44 per cent of the total. This is new.
Who gets the work permits? Most work permits go to a relatively small number of sectors of the economy and nationalities. The latest analysis, for the first six months of 2004, shows 30 per cent going to the health sector, 15 per cent to computer services and 11 per cent to hospitality. Although the private health sector accounts for some health worker permits, the majority go to the NHS. This concentration in a few sectors applies also to the nationalities of the workers concerned. Indians are the largest group, accounting for 29 per cent, up from only 8 per cent in 1995. Other nationalities which have become major suppliers through the work permit system are Filipinos, South Africans and Malaysians. In contrast, Americans, who took up a third of work permits in 1995, are down to only 11 per cent, while the proportion of Japanese is only a third of what it was.
Although the work permit scheme remains the principal route of entry for non-EU labour, several other possibilities exist. Some of these, such as those for working holidaymakers, seasonal agricultural workers, au pairs and domestic servants, are long-standing. Others, notably the sectors based scheme, worker registration scheme and highly skilled migrant programme, are recent.
Working holidaymakers, for example, are generally well educated and adaptable. Mostly they come from Australia, Canada, New Zealand and South Africa, which together accounted for 88 per cent of the total in 2003. The seasonal agricultural workers' scheme began after the second world war to facilitate the movement of young people across western Europe. Today it brings in young people mainly from eastern Europe, predominantly Ukraine, Bulgaria, Russia and Belarus.
Two new schemes are the sectors based scheme, designed to ease labour shortages in food processing and hospitality, and the highly skilled migrant programme. The most recent information suggests that Bangladeshis, Ukrainians and Pakistanis are the largest groups in the former. The latter scheme is different, since a permit is given to an individual rather than an employer.
Launched in January 2002, it is designed to help the UK compete globally by allowing individuals with exceptional skills and experience to come to the UK to seek and take work or self-employment. In the first half of 2004, 3,200 applications were approved; the largest group were Indians, followed by US citizens.
Asylum and family reunion
Two other significant categories of inward migrant are asylum seekers and those coming for family reunion.
The trend in numbers of asylum seekers has been something of a rollercoaster for the last 20 years (see right). From not many more than 4,000 in the mid-1980s, the number of asylum applications fluctuated in succeeding years before peaking at 84,130 in 2002 (103,000 including dependants). Since then, numbers have declined sharply, to 49,405 in 2003 (60,000 with dependants), the fall continuing in 2004. In the first nine months of 2004, there were 25,465 applications, down a third on the same period the year before, and only 12 per cent were accepted for some form of protection. It was suggested that the sharp fall in applications between 2002 and 2003 was the result of a government policy to divert potential asylum seekers into other, legal routes of entry. However, a national audit office study in 2004 found no evidence of diversion. Indeed, during this period most other European countries also experienced falls in asylum seeking. This was partly because of the reduction in the flow from Afghanistan and Iraq.
Much is made of the need to deport those immigrants, mainly failed asylum seekers, who do not have the right to stay, and, indeed, the number of people told to leave the country has increased, from 31,000 in 1997 to 64,000 in 2002. But there are no reliable data on how many people have actually left.
As a proportion of the total annual inflow of non-British people (the total was 407,000 in 2003), asylum is not as significant as recent media comment suggests. Nevertheless, for the period 1993—2003 the proportion of asylum seekers did rise. From 1999—2002 it hovered around a quarter of the non-British inflow, reaching a high of 27 per cent in 1999, before declining to just under 15 per cent in 2003.
Family migration takes two forms: reunion and formation. The first refers to family members coming to join a person already in residence; the second relates to the practice of seeking a marriage partner abroad. Both are distinct from family members accompanying labour or other migrants as dependants.
There are no statistical sources which give a clear indication of the scale of family migration. The statistics most often quoted relate to spouses and dependants (children and grandparents) granted settlement. In 2003, the number of these was 95,000, around double the figure in 1998, and they accounted for two thirds of all settlements.
Irregular migration
The irregular resident population consists of the foreigners living illegally here at any one moment. They may have entered illegally and remained so, or entered legally but become irregular because of a status change, such as overstaying the terms of their visa. The group differs from irregular entrants who entered the country illegally but may subsequently have had their position regularised, perhaps by being granted asylum. Irregular workers may be illegally resident and illegally working; others may be legally resident but become irregular by taking employment against the terms of their entry: examples include foreign students taking paid employment for more than their allotted hours.
To date no European government has produced an official estimate of its irregular population. This does not stop occasional media quotations of, usually large, round numbers best described as
guesses. The best information available comes either from amnesties or from border apprehensions. Although the UK has had several partial amnesties, usually associated with reducing the backlog of asylum applications and appeals, these have not been as big as those in some Mediterranean countries. Apprehension data provide an indicator of trends, but can never be the basis of a reliable estimate of irregular migrant flows, depending as they do on intelligence, spot checks and the resources the authorities make available for border control.
The costs and benefits of migration
Much of the debate about immigration focuses on whether or not it is to the benefit of the host population. No one seriously argues that there should be no immigration; the debate is about how much and what sort. Attempts to produce a balance sheet even at an aggregate level, let alone of the different types of movement, are notoriously difficult. In effect, protagonists try to weigh the partially measurable economic implications against the unmeasurable cultural and social ones. Even the factors we can measure in some way, such as fiscal or demographic ones, require assumptions that are easy to challenge.
Advocates of high levels of migration argue that migrants solve skill shortages and fill undesirable jobs, thus making the economy more competitive. Migration is also deemed to help us to counteract population ageing, either through revitalising the national population or providing people to pay pensions. It is also claimed that migrants make a net contribution to the national accounts by paying more in taxes than they consume in benefits. Migrants are also said to bring innumerable but unmeasurable cultural advantages.
All these things are difficult to demonstrate, principally because of a lack of data and the difficulty of allocating costs or benefits to different types of migrant: for example, the impact of a highly skilled migrant and an overseas undergraduate are not the same. Calculation of the economic effects of migration is usually done at the aggregate level, yet studies have demonstrated that the distribution effects of migration between regions, sectors and types of migrant are critical in any overall assessment.
Much attention is given to the need for migration as a means of replacing an ageing workforce. For the UK, this requirement is at least ten to 15 years in the future, when the baby boomers of the mid-1950s to mid-1970s start to retire; even then, increased female activity rates and higher productivity may provide alternative solutions. In any case, studies indicate that this approach would require sustained immigration rates at unacceptable levels, not least because migrants also age.
The fiscal benefits of migration have been much discussed in the literature in many countries. By and large they show immigration to be fiscally neutral; variations in calculations depend on which variables are included in the models and how much weight is given to them. A major concern in these models is the degree to which they consider the impact of migrants on the native workforce, particularly the least skilled end. When these are included, the fiscal impact of migration is more likely to be negative. The one study of the fiscal effect for the UK estimated a net annual gain to the economy of £2.5bn, but the authors admitted that in the absence of better data this could only be a rough approximation. Other authors have suggested the gain is much lower or non-existent.
Overall, it appears from the studies of the last quarter-century that there is no definitive answer to the question of how much immigration benefits the economy. Some types of migration will bring more benefits to the state than others, but it is not easy to say which, nor what their downside might be. Being able to import skills in short supply might reduce training of the native population, to the detriment of all. Conversely, recruiting a foreign management specialist might help native workers to be more efficient.
What of the future?
It is difficult to see immigrant numbers to Europe falling sharply—among other reasons, eastern Europe will join the west as a zone of net attraction. In recent years, international migration has become the main component of change in the UK population. The latest (2003) population projections from the government actuary's department assume that this will continue. Total population is projected to increase from 59.6m in 2003 to reach 65.7m by 2031, a rise of 6.1m. The natural change component (the difference between births and deaths) will add 2.506m, and net inward migration 3.64m. The projections assume a constant annual net immigration of 130,000. All that can be said of this assumption is that for the moment it is the best guess and that some figure has to be put into the model. Given recent net inflows into the country, 130,000 is reasonable but the longer-term historical trend is one of sharp fluctuation.
