Student visas and long-term immigration estimates: A simple relationship?

Last week (29 November 2012) the Office of National Statistics (ONS) released the new provisional long-term migration estimates, which show a decline in overall net migration to an estimated 183,000 for the year to March 2012, down from 242,000 for the previous year.

The decline was driven by both increasing emigration and falling immigration. Within the changes to immigration, the ONS identified falling numbers of people arriving to study as the primary reason for declining immigration.

When discussing migration statistics, it is important to keep in mind that the ONS’s official estimates of Long-Term International Migration (LTIM) are based on data from the International Passenger Survey (IPS) which defines migrants as people moving to the UK with the intention to stay for at least one year.

The LTIM data shows student immigration in the year to March 2012 falling by 19,000, from 232,000 to 213,000, which as the ONS pointed out, was the biggest single factor in the reductions to immigration over this period. Nonetheless, declines in other immigration routes were also in evidence: work immigration fell by almost as much (-17,000) and from a much lower starting point, and family migration also declined substantially (-10,000) over the same period.

But the decline in students is meaningful. Students are still by far the largest group of non-EU migrants to the UK, and numbers in the IPS estimates had tripled in a decade, from less than 60,000 in 2001 to more than 180,000 in 2010 (see Figure 1). The recent decline in student immigration follows a series of policies introduced by the coalition government to reduce overall net migration to the ‘tens of thousands’ by 2015.

Figure 1

Alongside official estimates of student immigration through March 2012, data released last week also show a continuing decline in student visas being issued, extending through September 2012. Interestingly, the decline seems to be concentrated outside the university sector. The best indication here comes from data on visa applicants based on Confirmations of Acceptance for Study (CAS’s). Comparing the last twelve months to the previous twelve (years ending March 2012 and March 2011), CAS’s increased by 1% (to 155,821) for those planning to study at university, while those in the further education sector, English language schools and independent schools fell by 67% (to 32,900), 76% (to 4,445) and 17% (to 14,087) respectively. Figure 2 shows quarterly CAS data since 2010 for the various types of educational institutions; the university (higher education) sector rises and falls seasonally but has changed little overall, whereas other types of institutions show downward trends.

Figure 2

Because overall student visas continued to fall through September 2012 (see Figure 2) whereas IPS data (the basis for official estimates of LTIM) are available only through March 2012, one might expect that IPS estimates of student immigration will continue to fall sharply. But it is difficult to predict exactly how IPS estimates will respond to changes in visas issued.

Figure 3 below shows data on student visas issued and estimates of non-EU student immigration from the IPS. (Since EU students do not require visas, it makes sense to compare visas to non-EU immigration only. The data in Figure 3 also exclude student visitor visas, as these do not grant permission to remain in the UK for long enough to qualify as a long-term migrant in the IPS data.)

Figure 3

As the figure shows, there are consistently more student visas than estimated student migrants based on IPS data in any given year. It also shows that trends in IPS immigration estimates broadly conform to trends in visas, but at times diverge as well.  The number of student visas issued annually declined by nearly 40,000 from 2009 to 2011, but over the same period, IPS estimates of non-EU migrants entering the country to study rose by about 25,000.

It is difficult to predict 2012 immigration estimates from 2011 visa data. In 2010, the Migration Advisory Committee (MAC) used past data to estimate the impact of reducing visas on various types of immigration in the IPS data – including student immigration (MAC report pages 292-294). As part of this process, the MAC used several years of data (2006-2009) to compare the number of visas issued with the estimates of the number of migrant inflows from the International Passenger Survey.

This created a ‘scaling factor’ for student visas of 0.54 during the period 2006-2009). In simple terms this means that each student visa issued translated into 0.54 non-EU student migrants in the IPS inflow data (or to put it even more simply, for every 100 visas issued, the IPS indicated that 54 migrants moved to the UK to study.)

This scaling factor changed in subsequent years. In the following year (2010) the estimated number of student migrant inflows in the IPS data was 0.71 of visas issued, and in 2011 it was an even higher proportion, 0.75. This narrowing of the gap between visas and immigration estimates is clear in Figure 3 above.

The reasons for the discrepancies between visas and immigration estimates are not fully understood, but we know that several factors are likely to be involved. The divergence may arise from the margin of error in survey-based IPS estimates, a proportion of visa-holders intending to stay for less than one year, and time lag between the issuing of a visa and the corresponding visa-holder arriving in the UK.

The data suggest that the discrepancies are not the straightforward result of time lags alone. Figure 3 shows that between December 2010 and December 2011 – a year after the decline in visas started – the number of non-EU student migrants to the UK did not show a significant decline in the IPS data and instead remained essentially unchanged (falling from 181,000 to 180,000), so the drop in visas issued in 2010 did not translate into fewer migrants in the 2011 IPS estimates.

But while understanding the causes may not be straightforward, it is clear that the discrepancy between the number of visas issued and the number of students arriving for more than a year (as estimated by the IPS data) is real, and that the size of this gap can fluctuate substantially. This means it has not yet been possible to create a predictive model that links changes in visa numbers to corresponding changes in number of student migrants. Declining visas most often mean a drop in IPS but the proportion is very difficult to predict. The only sure way to translate the last six months of visa data into IPS immigration estimates is to wait and see.

Related material

• MAC report – Limits on Migration: Limits on Tier 1 and Tier 2 for 2011/12 and supporting policies, Nov 2010
• Migration Observatory commentary – Entries, Exits, Errors