Price Sensitivity of Non-EU Students to
British Higher Education
Arnaud Chevalier
IZA (Bonn)
Xiaoxuan Jia
Anglia Ruskin University
Abstract
We build a panel of all incoming undergraduate students by country of origin at UK
universities to study their price sensitivity to British higher education. After controlling for
institutional quality and endogeneity of fees, we find that overall, overseas (non-EU) students
do not appear to react negatively to changes in tuition fees, but better ranked institutions,
and institutions that are popular among home/UK students attract more students from
outside the Europe.
JEL codes: F22, I23
Key words: higher education, tuition fees, price sensitivity
Acknowledgement: we would like to thank the Higher Education Statistics Agency (UK), and
John O’Leary (Times Good University Guide) for providing us with the student population data
and university rankings used in this analysis.
We thank Anna Vignoles and Colin Green for comments. An earlier version of this manuscript
was part of Jia’s PhD thesis.

Corresponding author: Xiaoxuan Jia, [email protected]
I.
Introduction
Previous literature on schooling and human capital accumulation have shown that
overall students are insensitive to university tuition fees (Heckman et al, 1998; Cameron and
Heckman, 1999; Canton and De Jong, 2005). It is the long term family factors rather than short
term credit constraint that affects students’ decision to invest in higher education. The focus
here is on international students, a group that may be more sensitive to price changes as they
are by definition more mobile. We rely on a panel of new entrants by country of domiciles at
British institutions, to assess their price sensitivity to university tuition fees over the period of
2002 to 2012. Tuition fees for home and EU students are regulated by the government but
institutions are free to set their own fees for students from outside Europe. We use the
variation in fees between and within institution to identify its effects on students’ registration.
Due to tight regulations of the local students market, UK institutions have expanded
their international outlook. Despite, an average fee of £12,000 for overseas students, one of
the highest in Europe1, the UK is the second largest destination countries for international
students and caters for 560,000 international students, or 13% of the total international
student population (OECD, 2013). Moreover, export of UK higher education is estimated to
be worth £10 billion in 2011/12 (HEFCE, 2014)2. Despite its importance to the economy, little
is known about the demand for higher education from international students, and even less
about their choice of institutions within a country. Unlike previous studies on the
determinants of international student migration, we contribute to the existing literature by
analysing the effect of institution specific tuition fees to assess how price (tuition fees)
sensitive overseas students are once they have decided to come and study in the UK. This
should be particularly relevant since unlike UK’s competitor countries such as Australia, the
UK government has tightened immigrations rules concerning international students over the
past decade. If international students are insensitive to tuition fees, such immigration policies
may pose a threat to institutional income and the UK economy in the long run.
Theoretically, individual study abroad when foreign education is associated with a
greater flow of income either abroad or back home, than local education, after accounting
1
Nominal figure, http://www.studyineurope.eu/tuition-fees
OECD report defines international students based on their country of citizenship, whereas HEFCE report, like
the current analysis, define international students as those coming from outside the EU.
2
for the costs differential in obtaining education (see Dustmann and Glitz, 2011, for a review).
In turn, studying abroad may affect an individual’s labour market outcomes either through
capturing the wage premium provided by host country’s labour market (Rosenzweig, 2006)
or obtaining higher returns back home, after gaining relevant skills and work experience
abroad. In this model, the decision to study abroad is determined by the costs of education
in both countries, the differences in the returns on skills in both countries, and the costs of
(return) migration—including non-financial costs such as family circumstances.
International students consider the relative costs, quality, and access to higher
education when making their decisions. While questionnaire evidence suggests that the
comparatively high cost of education in the US is a major obstacle to migration, behavior
seems to show otherwise. In a study of the determinants of student migration from 203
countries to 13 host countries, tuition fees had no significant effect on international
enrollment (Beine et al, 2012). This result is surprising considering that the empirical evidence
suggests that a $1,000 increase in annual tuition fees reduces the participation of US-born
students by 6–8 percentage points (Leslie and Brinkman, 1987)3.
Tuition fees may have no significant effect on the demand for higher education by
international students because tuition fees represent a relatively smaller share of the total
costs than they do for native students, especially when grants and other forms of aid are
available. In contrast, international students do care about the perceived quality of higher
education (such as the number of institutions in the top tiers of international rankings, Van
Bouwel and Veugelers, 2010).
We use undergraduate enrolment data for UK universities to assess the price sensitivity
of both the overall and the overseas student population, after controlling for institutional
quality. We first build a panel of first-year undergraduate students by institution, and country
of domiciles (6 domiciles overall, namely England, Wales, Scotland, Northern Ireland,
European (EEA) countries, and non-EU countries). We then assess the price sensitivity of nonEU students at the country specific level. Since a key concern of such analysis on tuition fees
is that the fee variable may not be strictly exogenous, we test the robustness of our findings
by using a two stage least squares estimator and instrument the fee variable with its one
period lags. Overall, we find that overseas students do not appear to react negatively to
3
Estimates for European students tend to be smaller.
changes in tuition fees, but better ranked institutions tend to attract more students from
outside the Europe.
The rest of the paper is organised as follows. Section II reviews previous literature.
Section III describes the institutional background of our analysis. Section IV provides details
of the data. Section V introduces the model. Section VI describes the results, and section VII
concludes.
II.
Literature review
Rosenzweig (2006) proposes two factors that drives the international migration of
students, a lack of educational resources/facilities and a low return to skills in the home
country. He finds that the US wage premium (over other countries) encourages foreign
students to migrate. Clemens (2013) finds that location alone can explain a significant part of
return to skills. De Vreyer et al. (2010) find that experience abroad results in a substantial
wage premium for migrants returning from an OECD country, but not from other countries.
Based on data of 18 European countries, Van Bouwel and Veugelers (2009) find that the
quality of higher education is positively linked to the number of students migrating from one
country to the other. Similar findings can also be found in Gordon and Jallade (1996). Quality
of education can also influence students’ decision to migrate within a country. Baryla and
Dotterweich (2011) analyse student migration in the US and find that colleges that provide a
recognised quality degree programme attracts more non local students. Using panel data on
the flow of international students to North America and a number of developed European
countries, Kahanec and Kralikova (2011) find that indeed, the quality of higher education in
the receiving country is an important factor in determining the inflow of foreign students, but
no significant correlation is found between tuition fees and size of overseas student
population. They contribute the latter finding to reverse causality in the sense that popular
universities can afford to charge a higher price. Similarly, Beine et. al. (2012) use data of 13
OECD countries to analyse factors that influence international students’ choice of destination
countries, and find significant network effect in the migration of students, other factors
include housing prices and reported quality of university, but not registration fees. Their initial
findings even suggested a positive relationship between tuition fees and the size of the flow
of international students. They propose two explanations for this apparent positive results.
First, tuition fees may act as a signal of quality. Secondly, popular destinations, and
universities with high reputation can charge higher fees without losing their applicant pool.
In the UK, Soo and Ellliott (2010) used overseas subject-level application data to UK
universities from the period of 2002 to 2007, and find that fees may impact overseas students’
application decisions, and this effect is non-linear.
For previous analysis on the impact of fees on domestic enrolment, Kane (1995) used
US data for the period of 1977 to 1993 to evaluate price sensitivity of students to tuition fees
at US public higher education institutions. He find that states with higher level of tuition fees
experience lower enrolment rates, with the effect vary between 3.5 percentage point for
public 2-year institutions, to 1.4 percentage points for 4-year institutions, for each $1,000
increase in fees. Using US data on American public universities (4-year institutions) from the
period of 1991 to 2007, Hemelt and Marcotte (2008) find that a $1000 increase in fees is
associated with a reduction in college enrolment by about 2.5 percent. For the effect of tuition
fees on higher education participation in the UK over the same period, Dearden et al. (2011)
constructed pseudo-panels by aggregating individual level data (over UK-domiciled students)
and find that, higher tuition fees are negatively linked to university participation rate, with
the effect measured at 3.9 percentage points decrease in university participation for every
£1,000 increase in tuition fees. Seven German states briefly introduced university tuition fees
in 2012, and Dwenger et al. (2012) find that the probability of students applying to university
fall by 2 percentage points if they live in a state with tuition fees. Similar findings using
German data can also be found in Hubner (2012).
III.
Institutional set-up
The higher education system for home and EU students was entirely state funded until
1998. A series of reforms introduced by the government in late 80s and the early 90s boosted
higher education participation rate in the UK, from 15% in 1989 to 42% in 2005 (Wyness,
2010). This brought considerable strains to public financing for higher education, and the
funding per full time students dropped from £9,000 in 1989 to £4,850 (Dearden et al., 2011).
A student loans system was implemented in 1990 to relieve pressure on public funding for
student maintenance and later on tuition costs. In addition, a means-tested up-front tuition
fees of £1,000 was introduced for those entering universities from 1998 onwards.
Tuition fees vary across the four constituent countries of the UK. Scottish universities
initially adopted the £1,000 means-tested fees, but it was abolished by the Scottish
Government in 2001, and replaced with the graduate endowment scheme 4, which is a fixed
amount of £2,000 (cash term) payable in April following graduation. This was abolished in
2007 5 , and subsequent cohorts of Scottish students studying at Scottish institutions paid
virtually zero fees6.
The 2004 Higher Education Act abolished up-front tuition fees and introduced a
deferred fee of £3,0007 to be implemented in 2006 for students domiciled in England and
Northern Ireland. The fee increase was delayed among Welsh universities to 2007 and a grant
was introduced for Welsh students and European students (studying in Wales) to cover the
amount of fee increase (Wyness, 2010).
Responding to recommendations made by the Browne Review of higher education
funding in 2010, the government announced that the maximum tuition fees English
universities can charge will be raised to £9,000 for home and European students from 2012
onwards. For the 2012 tuition fee reform, the Welsh government increased the grant to
Welsh students to £5,425, whilst the rest of the £9,000 can be covered by tuition fee loans.
This applies not only to Welsh students studying in Wales, but Welsh students who wish to
study elsewhere in the UK. Tuition fees in Northern Ireland changed very much in line with
tuition fees in England, except in 2012 it decided to freeze the tuition fees for students who
are domiciled and enrolled at higher education institutions within Northern Ireland8.
Students from continental Europe who are studying at each of the four constituent
countries of the United Kingdom are able to enjoy the same benefits as locally domiciled
students, but in general UK students who are domiciled in one of the constituent countries
4
http://www.scottish.parliament.uk/S1_Bills/Education%20(Graduate%20Endowment%20and%20Student%20
Support)%20(Scotland)%20Bill/b22s1pm.pdf
5
For all that graduate after April 2007.
6
The tuition fees for Scottish students studying at Scottish universities currently stands at £1,820, which is
paid for by the Scottish government. http://www.saas.gov.uk/full_time/ug/young/funding_available.htm.
7
“Deferred” in the sense that students can borrow at effectively zero interest rates to pay for their tuition
fees, and repay their loans after they have graduated and earn above a set threshold.
8
http://www.bbc.co.uk/news/uk-northern-ireland-14843377
and wish to study in another do not enjoy the same financial support towards fees as a locally
domiciled students.
Higher Education funding councils provide teaching grants to institutions based on the
number of home and EU students there are in each institution. Hence, prior to 2012,
institutions were limited to the number of home and EU students they can recruit. The
recruitment of overseas students are not restricted as their tuition costs are not funded by
the public. Hence, British universities increased their effort over time to attract students from
outside the Europe. The number of overseas students studying at UK universities increased
from 100,000 in 1994/95 to nearly 300,000 on 2009/10, and institutional income generated
by overseas students went up from £455 million to £2,580 million (Universities UK, 2011).
In contrast, UK immigration regulations concerning overseas students have been
toughened rather than relaxed over time. Application costs for entry clearance and leave to
remain have more than doubled during the period of analysis (UKCISA, 2014)9. A Post-Study
Work scheme which allowed overseas students who graduate from UK universities to stay in
the country for up to 2 years to gain work experience, was abolished in April 201210, and
replaced with a Graduate Entrepreneur visa. However, overseas students who wish to apply
for the Graduate Entrepreneur visa need to be endorsed by their university, and universities
are limited to how many graduates they can endorse per year11. As such, once they have
completed their study in the UK, most non-EU students will have no choice but to return to
their country of origin.
IV.
Data
The data originates from three sources, over the 11-year period of 2002 to 2012. Firstly,
we have the aggregate first-year undergraduate student population data by institution, fee
status (eligible vs not eligible to pay home fees) and domiciles (208 countries of domiciles),
from the UK Higher Education Statistical Agency (HESA). This dataset covers all UK higher
education institutions. We also have the tuition fees data by year, institutions and students’
9
Throughout this period, the UK government has also gradually increased language requirement for a student
visa, as well as the amount of (maintenance) funds the applicant need to have in their bank accounts
10
https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/261421/tier1poststudywor
kguidance1.pdf
11
https://www.york.ac.uk/students/support/international/after/working/
domiciles12. Finally, to control for institutional quality, we have the Times Good University
Rankings. The Times is a popular newspaper in the UK and publishes the longest running
national university rankings in the country since 1992.
There are three levels of undergraduate tuition fees in the UK in general, namely home
and EU student fees, overseas fees for classroom based courses and overseas fees for
laboratory based courses. The fee information is sometimes recorded as a range rather than
a single figure, under one of the three aforementioned broad categories. We transform them
by taking the mean to make them comparable across institutions. In a few cases the fee
information is missing for some institutions in certain years 13 , and we replace them by
interpolating linearly. We use the classroom based course fees for overseas students14, and
all fees are deflated to 2013 prices.
We model students’ enrolment decisions by using the HESA first-year undergraduate
student population data, which are recorded in December of each academic year15. Since the
level of tuition fees for home/UK and European students vary across the four constituent
countries of the UK, we assign and verify tuition fees according to the region of the institution,
students’ country of domiciles and their fees eligibility. Note that European students are
treated as locally domiciled students in each of the four constituent countries in the UK.
As described in institutional set-up, tuition fees in the UK vary by 6 domiciles overall,
namely England, Wales, Scotland, Northern Ireland, EEA countries, and non-EU countries.
Hence, we carry out our first set of regressions after summing the number of students in each
institution per year and collapse the data by the 6 domiciles.
During the period of 2002 to 2012, some universities have changed their names or
merged with other institutions to form new universities. As such, we cleaned up the names
of institutions so that they are consistent across years, as well as across the fees, rankings and
12
UK/EU and overseas, the latter includes two fee bands, namely classroom based and laboratory based
courses.
13
17 observations with missing fee information, two of which are missing undergraduate home/EU fees.
Reason for missing not explicitly given.
14
For consistency reasons, since not all university offer lab-based courses to students.
15
We define academic year as the September of each year to August in the following year. We cannot rule out
the possibility that some students may have dropped out between September (when the academic year starts)
and December (when the HESA data is recorded).
student population datasets. We treat institutions as separate before the merger and as a
new institution afterwards.
Overall, merging the student population data with the tuition fees and university
rankings data gives us 119 institutions over the 11-year period. However, we note that since
the 2012 tuition fees information is collected from the Complete University Guide16, there are
only 91 institutions in that year17. With the aforementioned merging of universities to form
new institutions, the created panel is as such unbalanced. Figure 1 and 2 show the trends in
undergraduate enrolment and tuition fees by country of domiciles.
English students take up the biggest proportion of first year student population, with
an average of 2508 students per institution per year (Table 1, Descriptive Statistics). This is
followed by Non-EU students, and students domiciled in Scotland. Students domiciled in
Northern Ireland form the smallest group with an average of 112 students per institution per
year. Non-EU students pay the highest tuition fees overall, with an average of £10,600 per
year. English students pay the highest tuition fees among home and EU students, with an
average of £3,200 per year. Note that average tuition fees for Scottish students won't be zero
as they will be treated as fee-paying students if they study outside Scotland. The mean level
of tuition fees for overseas students stayed fairly stable during the period of analysis, with the
variation increased somewhat in more recent years (Figure 3).
V.
Model
We begin by estimating whether the number of first-year undergraduate students
enrolled at institution i in year t, yit, is a function of the level of tuition fees xit, while
controlling for institutional quality Qit. The number of students is log transformed to give a
better fit of the model. The main equation depicting the relationship is
log (yit) = β1i + f1(Tt) + δ1* xit +λ1Qit + εit
16
(1)
The person who collected the tuition fees information for previous years, Mike Reddin, passed away in 2011,
and the Complete University Guide took over the responsibility from 2012 onwards.
17
See Appendix A1. Although the fee information for a number of universities is missing in 2012 due to that
they have merged with other institutions prior, for other institutions, there is no clear reasons (based on their
characteristics) as to why their fee information is not available in that year.
where βi is the institution specific fixed effect (to control for time-invariant, unobserved
institution specific factors), f(Tt) is a function of time, which includes year dummies, linear
trend, or institution specific trends, Qit is the institution quality measured by the Times
university ranking score, and εit is the random error term. The main coefficient of interest to
be estimated, δ1, represents the percentage change in the number first-year students for a
one thousand pound change in tuition fees. The standard errors are robust and clustered at the
institutional level to control for within-institution correlations and heteroskedasticity.
Next we drop the observations of UK and EU students, and we assess overseas students’
price sensitivity to UK higher education as a function of tuition fees, quality of the institution,
and a set of country specific economic variables, g(Ect) for non-European country c in year t,
to control for the return and other financial costs of their education. The set of economic
variables include inflation rate (measured by the GDP deflator), GDP growth and year-on-year
exchange rate fluctuation (domestic currency to British pounds). Zit is the number of overseas
students at institution i in year t.
Zit = β2i + f2(Tt) + δ2* xit +g(Ect) + λ2Qit + uit
(2)
It is important to note that we implicitly assume overseas students have decided to
study in the UK, and we only assess their price sensitivity to tuition fees conditional on
studying in the UK. We replaced any missing observations with 0 to create a continuous time
series for each institution-country group. One of the concerns is that the fee variable may not
be strictly exogenous, since popular UK universities that attracts higher number of overseas
students can also afford to charge them higher fees. Hence, we test the robustness of our
results by using a two stage least squares estimator and instrument the fee variable with its
one period lag.
Table 1 shows the descriptive statistics for key variables, and we can see that among all
UK and European students, the English domiciled students pay the highest average tuition
fees among home and EU students during the period of 2002 to 2012, while the average
tuition fees for overseas students is around £11,000 (2013 prices). Figure 1 shows the number
of first-year undergraduate students by country of domiciles from 2002/03 to 2012/13. The
number of first-year English domiciled students, European students and non-EU students
have been gradually increasing until 2012/13, whilst the number of Northern Irish students,
Scottish students, and Welsh students remained relatively unchanged during this period.
Figure 2 shows the average UK tuition fees each year for students domiciled in different
regions. Tuition fees for non-EU students have been steadily increasing during the 11-year
period, from £10,000 to £12,000. English domiciled students have seen the highest increase
in tuition fees, whist Welsh and Scottish students have experienced lower fees, since they
have benefited from tuition fee grants and zero tuition fees if they choose to study locally.
Average fees for European student and Northern Irish students changed very much in line
with English students.
VI.
Results
We test the price sensitivity of the whole first year undergraduate student population
under the basic specification (equation (1)) in table 2. Overall results suggest increase in fees
lowers enrolment, but estimates under year dummies or the linear trend can be biased as
they do not take into account of the unobserved factors at the institution level. We find that
an increase of £1,000 in tuition fees lowers first-year undergraduate enrolment by 3.6% in
our favoured specification (column 3). We also test the heterogeneity in students’ sensitivity
to tuition fees by their country of domiciles. For doing so, we interact the fees variable with
domiciles dummies, and the results are presented in column 4. For English students (the
baseline group), every £1,000 increase in fees is associated with a 4% drop in enrolment. The
size of the estimate is consistent with the ones found in Dearden et al (2011) and Hemelt and
Marcotte (2008). The impact of fees more than doubles with Scottish and Welsh students.
Consistent with findings in Beine et al (2012), non-EU students appear to be rather insensitive
to changes in fees compared with the baseline group, whilst for students from European
countries, a £1,000 increase in fees is associated with roughly 2% increase in enrolment.
In terms of the effect of institutional quality on students’ enrolment decisions, we find
that an increase of 100 points in Times university-level ranking score increases the size of
overall first-year student population by about 2% in our favoured specification. The sign of
the coefficient is consistent with previous findings (Soo and Elliot, 2010) on the effect of
university rankings, which suggest better ranked institutions attracts more students, but the
results are not statistically significant under OLS fixed effects estimation.
Instrumenting the overseas fees with lagged fees under the two stage least squares
estimation (column 5) shows that a £1,000 increase in fees lowers the number of students by
about 0.05%. This estimate is very small in size and it is not statistically significant. However,
a 100-point increase in the Times ranking score increases the number of first-year students
by about 3%, and this effect is statistically significant at 10%.
We then assess the effect of tuition fees using the more detailed, country specific
overseas student population data, where we control for country specific fixed effects,
institutional quality, and country specific economic conditions (equation (2), Table 3). Overall,
we do not find statistical significant evidence that overseas students react negatively to
changes in tuition fees, but better ranked institutions attract more students, and higher GDP
growth and smaller exchange rate fluctuations in the home country increases the size of the
overseas student population at an UK institution.
VII. Conclusion
Estimating the price sensitivity of non-EU students to UK (overseas) tuition fees over
the period of 2002/03 to 20012/13, we find that overall, overseas students do not appear to
react negatively to changes in tuition fees, but better ranked institutions, and preferable
economic condition in the home country help to bring in more students from outside the
Europe.
Assessing the price sensitivity of non-EU students to UK (overseas) tuition fees has
important policy implications. Considering that average within-institution variation in the
number of overseas students is around 103, and the average tuition fees per student per year
is around £12,000 in 2012/13, this means effectively a change of £1.2 million in institutions’
income. However, contrary to the more open immigration policies in other popular
destination countries, where efforts have been made to allow international students to stay
and gain work experience after their study, the UK government on the other hand withdrew
such Post Study Work schemes, includes international students in the annual immigration
target, and gradually increased the entry cost for non-EU students to come and study in the
country. HEFCE (2014) reports that the drop in the number of non-EU students to English
institutions in 2012/13 is the first drop in 29 years. Hence, such policy changes could
potentially harm the financial wellbeing of UK universities and damage UK universities’
presence in the international higher education market.
Reference
Baryla, E. A. JR., and Dotterweich, D (2001) Student migration: do significant factors vary by
region? Education Economics, 9(3), 269-280
Beine, M., Noel, R., and Ragot, L.(2012) The determinants of international mobility of students,
CESifo working papers, No. 3848
Cameron, S. V., and Heckman, J. J. (1999) The dynamics of educational attainment for Blacks,
Hispanics, and Whites. National Bureau of Economic Research Working Paper, No. 7249
Canton, E., and De Jong, F. (2005) The demand for higher education in The Netherlands, 1950
– 1999, Economics of Education Review, 24(6), 651 – 663
Dearden, L., Fitzsimons, E., Wyness, G. (2011) The impact of tuition fees and support on
university participation in the UK, Centre for the Economics of Education, London School of
Economics
De Vreyer, P., Guber, F., and Robilliard, A. (2010) Are there returns to migration experience?
An empirical analysis using data on return migrants and non-migrants in West Africa,
Migration and Development, 307-328
Dustmann, C., and Glitz, A. (2011) Handbook of the Economics of Education, Volumn 4, 327 –
439
Gordon, J., and Jallade, J (1996) ‘Spontaneous’ Student Mobility in the European Union: a
statistical survey. European Journal of Education, 31 (2), 133 – 151
Harris, R. (1995) Overseas students in the United Kingdom university system. Higher
Education, 29(2), 77-92
Heckman, J. J., Lochner, L., and Taber, C. (1998) Explaining Rising Wage Inequality:
Explorations with a Dynamic General Equilibrium Model of Labor Earnings with
Heterogeneous Agents. Review of Economic Dynamics, 1(1) 1 – 58
HEFCE (2014) Global demand for English higher education: An analysis of international
student entry to English higher education courses. April 2014, 2014/08a
Hemelt, S. W., and Marcotte, D. E. (2008) Rising Tuition and Enrollment in Public Higher
Education, IZA Discussion Paper 3827
Kahanec, M., and Kralikova, R. (2011) Pulls of International Student Mobility, IZA Discussion
Paper 6233
Kane, T. J. (1995) Rising Public College Tuition and College Entry: How Well Do Public
Subsidies Promote Access to College? National Bureau of Economic Research, Working
Paper 5164
OECD (2013) Education at a Glance 2013, OECD Indicators
Rosenzweig, M. (2006) Global wage differences and international student flows. Brookings
Trade Forum, 57-96
Soo, K. T., and Elliott, C. (2010) Does price matter? Overseas students in UK higher education.
Economics of Education Review, 29, 553-565
Universities UK. (2011) Higher education in facts and figures
http://www.universitiesuk.ac.uk/Publications/Documents/HigherEducationInFactsAndFigur
esSummer2011.pdf
Van Bouwel, L., and Veugelers, R. (2010) Does university quality drive international students
flows? Department of managerial economics, strategy and innovations, Catholic University
of Leuven
Wyness, G. (2010) Policy Changes in UK Higher Education Funding, 1963-2009. Institute of
Education, Department of Quantitative Social Science (DoQSS) Working paper
Universities UK (2011) Higher Education in Facts and Figures.
http://www.universitiesuk.ac.uk/highereducation/Pages/HigherEducationInFactsAndFigures
2011.aspx#.VX9BtGDbL4g
Universities UK (2013) The funding environment for universities.
http://www.universitiesuk.ac.uk/highereducation/Documents/2013/FundingEnvironmentFo
rUniversities.pdf
UKCISA (2014) 5 years of Tier 4: key developments and changes. Special briefing.
Figure 1: Number of first-year undergraduate students by country of domiciles (in thousands)
No. of students
domiciled in the
rest of UK, EU and
non-EU
No. of students
domiciled in
England
350
40
300
35
30
250
25
200
20
150
15
100
10
50
5
0
0
200203 200304 200405 200506 200607 200708 200809 200910 201011 201112 201213
England
European coun
Non-EU
N.Ireland
Scotland
Wales
Note: source: HESA first-year undergraduate student population data (2002/03 – 2012/13)
Figure 2: UK average tuition fees for students of different country of domiciles.
Home and EU
tuition fees
Non-EU tuition fees
14
10
9
12
8
10
7
6
8
5
6
4
3
4
2
2
1
0
0
200203 200304 200405 200506 200607 200708 200809 200910 201011 201112 201213
Non EU
England
European Countries
N.Ireland
Scotland
Note: Source: Reddin (& Complete University Guide) tuition fees data (2002/03 – 2012/13)
Wales
5
10
15
20
25
Figure 3: Variation in UK university overseas tuition fees, 2002 – 2012
200203 200304 200405 200506 200607 200708 200809 200910 201011 201112 201213
Note: source: Reddin and the Complete University Guide tuition fees data (2002/03 – 2012/13)
Table 1 Descriptive Statistics
Variable
number of first year students
(by institution, academic year, and
country of domicile)
By students' domiciles:
Mean
570.47
England 2508.29
Wales 141.58
Scotland 220.66
Northern Ireland 111.91
European countries 148.77
Non-EU countries 274.62
tuition fees (2013 prices, in thousands,
£)
4.21
By students' domiciles:
England
Wales
Scotland
Northern Ireland
European countries
Non EU countries
Std. Dev.
162.26
Observations
6720
367.64
52.62
52.68
39.75
64.51
103.24
1128
1110
1111
1118
1128
1125
1.54
6712
3.16
1.84
1124
2.72
0.95
1110
2.86
1.76
1111
3.15
1.83
1114
2.80
1.70
1128
10.55
0.84
1125
ranking score
588.37
79.24
6720
Note: we report the within institutions standard deviations. Although Scottish
students pay zero fees in Scotland, they are subject to higher fees if they wish
to study in other parts of the UK. Fees for 2012 cohort of English and Northern
Irish students studying at Robert Gordon, Dundee, Glasgow, and St Andrews
(all Scottish institutions) are not available.
Table 2 Fee sensitivity of ALL students, controlling for institution quality
log (students)
(1) Fixed effects
VARIABLES
fees
times ranking score
-0.0236***
(0.00754)
0.0166
(0.0232)
-0.0387***
(0.00511)
-0.00648
(0.0138)
-0.0357***
(0.00803)
0.0199
(0.0157)
-0.0410***
(0.00896)
0.0187
(0.0157)
-0.0005
(0.0420)
0.0314*
(0.0173)
x
x
0.0598***
(0.00950)
0.0431**
(0.0211)
0.0204**
(0.00868)
-0.0581***
(0.0120)
-0.0751***
(0.0159)
x
x
x
x
4.250***
(0.102)
4.215***
(0.109)
European countries x
fees
Non EU countries x fees
Northern Ireland x fees
Scotland x fees
Wales x fees
institution specific trends
year dummies
linear trend
Constant
x
(2) 2SLS
x
4.284***
(0.154)
-77.02***
(10.45)
Observations
6,712
6,712
6,712
6,712
5,987
Number of groups
711
711
711
711
682
Adjusted R-squared
0.069
0.065
0.142
0.176
0.017
Note: "Number of groups" refer to observations by institution and domicile groups, namely
England, Wales, Northern Ireland, Scotland, European countries and Non EU countries.
English students is the baseline group in column 4. Robust standard errors in parentheses.
*** p<0.01, ** p<0.05, * p<0.1
Table 3 Fee sensitivity of non-EU students, controlling for country-specific
economic variables
VARIABLES
fees
times ranking score
inflation (GDP deflator)
GDP growth
exchange rate fluctuation
Non-EU students
(1) Fixed effects
0.328*
(0.170)
1.015***
(0.315)
0.0071*
(0.00377)
0.0233***
(0.00631)
-2.78e-06***
(3.75e-07)
institution specific trends
year dummies
linear trend
Constant
0.297*
(0.161)
0.959***
(0.188)
0.0030
(0.00357)
-0.0081
(0.00630)
-3.05e-06***
(3.49e-07)
x
-0.109
(0.0948)
0.444**
(0.188)
0.0077*
(0.00404)
0.0229***
(0.00669)
-2.88e-06***
(4.00e-07)
x
x
(2) 2SLS
-1.606
(1.521)
0.672**
(0.272)
0.0076*
(0.00426)
0.0208***
(0.00720)
-3.06e-06***
(4.48e-07)
x
x
x
-6.380**
(3.177)
-1,045***
(169.9)
0.985
(1.524)
Observations
53,067
53,067
53,067
45,961
Number of groups
9,768
9,768
9,768
6,586
Adjusted R-squared
0.021
0.018
0.045
-0.118
Note: "Number of groups" refers to observations by institution and (non EU) country of domiciles.
All except non-EU students are deleted (hence, there are only 11 observations for each
institution in the dataset, from 2002 – 2012). Robust standard errors in parentheses. *** p<0.01,
** p<0.05, * p<0.1
Appendix
Table A1 missing institutions in 2012
Institutions
Anglia Ruskin
Brunel
Chichester
City
De Monfort
Notes
Post 1992 new
university
University status Oct
2005
Post 1992 new
university
Edge Hill
Glamorgan
Gloucestershire
Imperial College London
Kingston
Lampeter
Lincoln
Liverpool John Moores
London South Bank
Loughborough
Middlesex
Newport
Paisley
Teesside
Trinity Saint David
University status May
2006
Post 1992 new
university-merged into
University of South
Wales April 2013
University status 2001
Post 1992 new
university
Merged with Trinity
University College in
2010 to form University
of Wales, Trinity Saint
David
Post 1992 new
university
Post 1992 new
university
Post 1992 new
university
Merged into University
of South Wales April
2013
Merged with Bell
College to form
University of West of
Scotland, August 2007
Post 1992 new
university
UMIST (University of Manchester
Institute of Science and Technology)
Merged with Victoria
University of
Manchester to form
University of
Manchester in 2004
Victoria Manchester
Merged with UMIST to
form University of
Manchester in 2004
Warwick
West of Scotland
Westminster
Wolverhampton
York St John
Post 1992 new
university
Post 1992 new
university
University status 2006