The Conservative party’s goal of reducing net migration to the tens of thousands was a driving force behind a series of measures from 2010 onwards restricting each of the three main routes through which non-EU nationals can apply to come to the UK: work, family and study. These policies have generated ongoing debate about how the UK should prioritize different types of migration and who should bear the brunt of reductions in eligibility for visas or settlement.
One strand of this discussion has been the role of international students in the immigration system. Students who arrive or leave for more than 12 months are counted in the same way as those who come and go for other reasons. Some commentators have argued that this should not be the case: that students should not be seen as contributors to net migration and that there should not be a policy target to reduce their numbers.
This analysis looks where students fit in the immigration system, what we know about how they contribute to net migration, and what it would mean to ‘take them out’ of the net migration target.
What are the reasons for and against excluding students?
It’s important to distinguish between the net migration statistics and the net migration target. The net migration statistics themselves are collected by the Office of National Statistics based mainly on a survey of passengers at UK ports and airports. While there are many ways of determining who counts as a migrant, the net migration statistics are based on an internationally recognized definition of people who move for 12 months or more. Taking students out of the statistics as they are collected by ONS is thus not an option unless ONS were to adopt a completely different approach to deciding who counts as a migrant.
However, this does not mean that students must remain in the net migration target. The target is a political goal that can be designed in any way the government sees fit – including or excluding people who arrive as students.
What are the arguments for and against including students in any target to reduce net migration? Three main arguments are typically advanced in favour of taking students out of the target. The first is that students are temporary, arriving for a specific course of study and then leaving after a few years – and therefore should not be treated in the same way as people who are more likely to settle permanently.
The second is that the public does not consider students to be ‘immigrants’ and are less likely to support measures to reduce numbers of students than other groups. Migration Observatory research in 2011 found that only 29% of the public included students among the groups that they think about when they think about immigrants, much less than the other major groups (asylum applicants, workers and family migrants). Students were also the group that respondents were least likely to favour reducing in number.
The third argument is that students bring economic benefits by paying tuition fees and spending money in the UK, and thus to reduce their numbers would be economically damaging.
By contrast, two main arguments are typically advanced in favour of keeping students in the net migration target. The first is that residents who are temporary still contribute in the short run to demand for housing, transportation, and other services that may become congested. The second argument is that some students are not temporary, and instead stay on to work and settle permanently in the UK.
The temporariness of international students is a major question when it comes to the debate about taking students out of the net migration target. The rest of this analysis examines what we know about whether students are temporary and what it would mean to take students out of the target.
It does not address the question whether ensuring that students go home rather than staying on to work in the UK is a desirable outcome, as this is part of a much larger debate about the potential benefits and costs of skilled migration more generally. It also does not take a position on the question whether reducing net migration overall is a desirable outcome, and whether there should be quantitative metrics for doing so. As we have said before, there is no ‘optimal’ level of migration and all decisions on how generous or restrictive policies are involve trade-offs.
How temporary are international students?
Student visas allow non-EU nationals to come to the UK for a specific course of study with a licensed education provider. They can be issued for a range of different courses, such as university degrees, further education courses and English language training. The visas are temporary and expire shortly after the course ends. However, student visa holders may be able to stay legally in the UK if they switch to another visa category, such as work or family.
Switching from study to work has become harder in the past few years because of the elimination of the post-study work programme, which previously allowed students to work after graduation. It provided an easier transition into the labour market than the Tier 2 (general) visa, since former students did not have to meet the salary requirements of Tier 2 visas or find an employer willing to meet the responsibilities of sponsorship.
Some policies facilitating students’ transition into the labour market after graduation using Tier 2 visas remain in place. People switching from study to work are not subject to the cap on Tier 2 sponsorships, which means that they currently do not face newly increased salary requirements that have kicked in now that the cap has been met for the first time. Their employers are also exempt from the requirement to show they have looked for UK or EEA candidates.
The number of people switching from study into other categories fell substantially between 2011-2012 and 2013-2014 (Figure 1). In 2014, just under 12,000 people extended their stay in the UK by switching from study to another category. The majority (59%) switched into work, while 33% switched into the family category.
The annual number of students granted entry visas to the UK (as main applicants) has fluctuated around 200,000 over the past few years, after peaking at 273,000 in 2009. In other words, the vast majority of people who enter on student visas are not switching into other categories.
Data from the International Passenger Survey (IPS) – which is a sample survey used to estimate net migration – provide a different angle on the movements of international students. The IPS uses different definitions from the visa data, and so is not directly comparable. Most importantly, it only includes people who intend to move for 12 months or more, while the visa data include some people who move for less than a year, as well as any people who are granted visas but change their plans and never come to the UK. As a result, we should expect the visa data to include more people than the IPS data.
The IPS asks respondents why they are coming to the UK, allowing us to identify students on their way in. Since 2012, respondents who are surveyed on the way out of the UK are also asked why they originally came. This allows an estimate of the number of people who come as students and – eventually – the number of them that leave.
The IPS suggests that 135,000 non-EU students entered the UK for study in 2014, plus or minus 17,000 due to statistical error (meaning that we can be 95% confident that the true number lies between 118,000 and 152,000.) The number of people who had previously arrived as students and who are estimated to have emigrated from the UK in 2014 was 44,000, plus or minus a margin of error of 6,000.
Taken together, this suggests that in 2014, net migration of students according to the IPS was 91,000 – that is, 91,000 more new students arrived than former students left. This snapshot must be interpreted carefully, because the people arriving and leaving are part of different cohorts. We do not yet know how many of the 2014 student cohort will leave, as many are not expected to do so for a few years. (In 2013, for example, the IPS estimates that about one quarter of former study migrants of all citizenships leaving the UK had arrived the previous year, and a further half had arrived between 2009 and 2011 inclusive.)
So while the outflow of former students was about 33% of the inflow in 2014, we cannot say based on this snapshot alone that ‘33% of students go home’, because people arriving and departing are from different cohorts. Eventually, it will be possible based on IPS data to construct an estimate of what share of each year’s cohort of students leaves, but this is not currently possible because the relevant data only goes back to 2012. However, if the current number of student inflows and outflows remained stable at these levels for several years, it would suggest that a majority of students were not going home.
With more years of data we will eventually get a better sense of whether outflows are changing over time and whether they have responded to recently introduced policies. What is clear for the moment, however, is that for the past three years the estimated inflows of students has been significantly higher than the number of self-reported former students estimated to be leaving (Figure 2).
Why does the IPS suggest that a significant number of students remain in the UK, while the visa data suggest that a relatively small number switch from study to other categories within the country?
There are different possible explanations for this, and unfortunately currently available data do not allow us to know which bears more responsibility. One possibility is that the IPS is not capturing the outflows of former students accurately and thus understates emigration of former students. Survey respondents must recall their initial reason for coming to the country a few years earlier. Since mixed motivations for migration are common, it is possible that respondents may have been systematically less likely to say they arrived as a student when leaving – especially if they also worked for a couple of years after graduation. Others may be planning to return in less than 12 months, and thus not be counted as emigrants. In addition, the IPS survey itself is imperfect for various conceptual and operational reasons, although ONS has made significant efforts to improve the quality of the data within its budget constraints, and continues to review its accuracy. Finally, IPS estimates have relatively large margins of error, as shown in Figure 2, so it is also possible that at least some of the difference between student inflows and outflows is the result of statistical noise.
The second explanation is that students are overstaying their visas rather than leaving the UK when their leave to be here has expired. The ongoing programme to collect exit records for people leaving the UK could, in theory, help ONS to understand how much this is happening. However, for various technical reasons it remains unclear whether and when the data collected from exit checks will be of high enough quality to get an accurate picture of overstay rates.
For now, perhaps the safest conclusion is that the temporariness of international students remains uncertain.
Temporariness and the bounce effect
Whether international students are temporary is the key factor governing their impact on net migration in the long run.
If 100% of students left the country within a few years, then over the long run they would not contribute to net migration, even under the current statistical measures. This is because students would add to immigration numbers when they arrived and add to emigration numbers when they left, with a net impact of zero over time. This would be the case over the long term regardless of whether inflows of students were 10,000 or 200,000.
Even if all students were temporary they might affect total net migration in the short run, however. This is because a sharp increase or decrease in student inflows in one year would not be reflected in higher or lower emigration until a few years later, when that cohort of students left the country. Elsewhere we have referred to this relationship as the net migration ‘bounce effect‘. It means that even if all students were temporary, an increase in the number of student arrivals would make it harder for the government to meet the net migration target in the short run, although not in the long run. The strength of the bounce effect for students or any other group will depend on what share of them are in fact temporary.
Taking students out of the target: How would it be done in practice?
If the government retains the IPS as the main metric for setting quantitative goals on migration, taking students out of the target would essentially involve calculating a new level of non-student net migration from the IPS and using it as an alternative metric against which policy is assessed.
The most obvious way to do this is to calculate net migration of all people who do not arrive as students, by removing students from inflows and removing former students from outflows.
How would this affect the targeted level of net migration? Table 1 shows the various components of net migration and allows us to calculate what net migration would be without students. It includes the calculation for students of all citizenships (the left hand columns) and for students who are non-EU nationals (the right hand columns).
The standard measure of net migration, including students, is total inflows (i.e. line 1 of Table 1) minus outflows (line 2). For 2014, this means 592,000 – 300,000 = 292,000. Note that this is slightly different from the 318,000 figure widely reported in the May 2015 immigration statistics, which is known as the estimate of Long Term International Migration (LTIM). LTIM is based on the IPS but also includes various adjustments to the IPS (for example to account for asylum applicants). However, LTIM does not include a breakdown of the number of former foreign students leaving the UK, so for the purposes of this calculation we use the slightly lower IPS figures.
If we take total non-student immigration (line 7) minus total non-student emigration (line 8), we get 404,000 – 235,000 = 169,000 in 2014. This is 123,000 lower than the net migration figure that includes students (line 6). If only non-EU students are removed, the estimate of non-student net migration is slightly higher, at 201,000 – a reduction of 91,000. Effectively, removing students from the net migration calculation reduces the headline figure by exactly the estimate of student net migration in a given year.
In other words, a major reason non-student net migration is likely to be lower than overall net migration is that not all students are temporary. If all students were temporary then the effects of removing them from the target would be considerably smaller.
Table 1: inflows and outflows of students and other groups, 2012-2014, all citizenships (thousands)
|Removing students of all citizenships||Removing non-EU students only|
|3||Net migration from IPS||164||190||292||164||190||292|
|5||Outflow former students||67||72||65||49||50||44|
|6||Difference between inflow and outflow||108||99||123||90||72||91|
|7||Total inflows without students||287||314||404||323||363||457|
|8||Total outflows without students||231||223||235||249||245||256|
|9||Net non-student migration||56||91||169||74||118||201|
Source: ONS International Passenger Survey
In summary, taking students out of the net migration target would have two main effects: (1) removing the short-term impacts of any student net migration ‘bounce’ (for example due to a sharp change in inflows or outflows in a particular year) and (2) bringing the government closer to meeting a given target (such as 100,000) by exempting some long-term migrants from it.
The bigger questions
These calculations all assume that the IPS-based net migration statistics remain the primary metric for assessing the government’s goal of reducing net migration, and that everything else about the target would remain the same.
However, the discussion about students is in some respects just one part of a broader question about how to construct policy targets that reflect the government’s policy priorities. Other aspects of the methodology for assessing the effects of policies could also be considered – such as the fact that the net migration target includes British and EU citizens, who do not need to apply for visas and whose movements are thus more difficult to influence.
There are several different ways to measure the scale and composition of migration, including the long-term increase in the size of the foreign-national or foreign-born populations living in the UK, long-term immigration inflows, long-term net migration, or the number of long- or short-term visas issued to non-EU nationals in different categories. Other significant outcomes, such as the labour market integration of migrants after they arrive can also be measured statistically. These too could be explored as metrics to guide the policymaking process.
If quantitative metrics are to be prominent in the policy debate, it is worth considering how a suite of different measures, rather than just a single metric, could paint a more nuanced picture of migration in the UK that reflects the inevitably complex set of economic, social and political goals the government wishes to achieve.