Contra in Sts

Studies in Higher EducationVol. 33, No. 4, August 2008, 405–429

ISSN 0307-5079 print/ISSN 1470-174X online© 2008 Society for Research into Higher EducationDOI: 10.1080/03075070802211802http://www.informaworld.com

Does the fear of debt constrain choice of university and subject of study?

Claire Callendera* and Jonathan Jacksonb

aBirkbeck, University of London, UK; bLondon School of Economics, UKTaylor and FrancisCSHE_A_321347.sgm10.1080/03075070802211802Studies in Higher Education0307-5079 (print)/1470-174X (online)Original Article2008Taylor & Francis3340000002008Prof. [email protected]

The new student funding regime introduced by the 2004 Higher Education Act inEngland is predicated on the accumulation of student debt. Variable tuition fees, repaidby student loans, will increase average student loan debt on graduation. This articleexamines how fear of debt and financial constraints affect prospective students’ choicesof where and what to study. Using data derived from a survey of about 2000 prospectivestudents, it shows that financial issues constrain lower social class students’ choice ofuniversity far more than those from other social classes. It demonstrates that fear of debtis related to two key financially-driven decisions – applying to a university with lowliving costs, and applying to one with good term-time employment opportunities – butonly for students from low-income families. However, concerns about debt do notinfluence their choice of qualification and subject. The article concludes that low-incomestudents are more likely than their wealthier peers to perceive the costs of highereducation as a debt rather than an investment.

Introduction

The 2004 Higher Education Act heralds a radical shift in England’s higher education fund-ing, and particularly in student finances. The Act, which came into force in 2006/2007, hasderegulated undergraduate tuition fees and introduced a quasi-market in higher education.Universities in 2006/2007 can charge up to a maximum of £3000 for any undergraduatecourse. However, all students, irrespective of their family’s income, now pay tuition fees.So, the means-tested flat rate tuition fees paid up front introduced in 1998 have been replacedwith deferred fees repaid after graduation via an optional student loan. Consequently,students now can take out a loan for both their living costs and their tuition fees. And it isanticipated that the take up of loans for tuition will be the same as the take up of loans formaintenance, around 80%. Thus, the new student funding system is increasingly predicatedon the accumulation of debt.

Concerns about the effect of the proposed changes on widening participation, socialmobility and social class differentials prompted various schemes specifically aimed atsupporting students from low-income backgrounds. First, a means-tested grant of up to £2700was re-introduced. Full grants are paid to students from households with incomes of £17,500or less, which represents about 85% of Britain’s median household income, and from 2008/2009 this income threshold will rise to £25,000. Second, student debt is to be written offafter 25 years, which is an important safety net for low earning graduates, and especiallyfor women with interrupted patterns of labour market participation due to childbearing andrearing. Finally, universities charging the maximum tuition fees must give low-income

*Corresponding author. Email: [email protected]

406 C. Callender and J. Jackson

students non-repayable bursaries of £300 a year to supplement their state-funded grants andmaintenance loans. On top of this mandatory minimum, the government has also encourageduniversities to provide additional discretionary financial support to promote widening partic-ipation.

It is far too early to assess the impact of the 2004 Higher Education Act on the highereducation sector, or on students, as it only came into force in 2006/2007. Despite the newgrants and mandatory bursaries for low-income students, however, there is little doubt thatthe introduction of variable tuition fees, paid for via an additional student loan, will increaseaverage debt on graduation (and the length of time it takes students to pay off these debts).Estimates of this rise vary substantially. For instance, the government suggests that debt willincrease to £15,000 for the first cohort of fee payers (Department for Education and Skills2003c), while Blanden and Fitzsimons (2007) believe it will be higher, in the region of£20,000. This will be double the average level of debt on graduation previously for the rich-est students and one and a half times higher for the poorest students. To put this sum inperspective, it means that around a half of students will be borrowing more money than theirfamilies’ annual income. Rising debt is also likely to change the relationship between thecosts and benefits of higher education, although the buoyancy of the graduate labour marketwill mediate this.

The coming years, therefore, will see much interest in the intended and unintendedconsequences of these reforms. Of particular interest will be their impact on existingsocial class inequalities in patterns of participation, including choice of university andcourse. This is the focus of this article. It aims to examine the potential impact of thereforms on students. Specifically, it concentrates on attitudes to debt and the costs ofhigher education. Our contention is that if concerns about debt and the costs of highereducation proved to be an issue for potential students, at a time when both debt andcosts were relatively low (the time of our survey), then, all things being equal, they arelikely to be even more of an issue once both rise as a result of the 2004 Higher Educa-tion Act.

This article is based on data derived from a national survey of just under 2000 final yearstudents in further education colleges and school sixth forms. The study investigatedprospective higher education students’ attitudes to debt, higher education and studentsupport, and their impact on participation and choices (Callender 2003). We have alreadyreported that debt aversion deterred people from the lower social classes from applying touniversity, but not middle-class or upper-class students (Callender and Jackson 2005). Adeterrent effect held even after controlling for students’ educational ability, their attitudestowards the benefits of going to university, their cultural beliefs, and a range of socio-economic factors. There is, therefore, clear evidence that students from lower-incomefamilies are more sensitive to the costs of higher education than students from wealthierbackgrounds. They are thus more likely to see university in terms of unacceptable debtaccrual rather than a beneficial investment.

Yet to understand the full impact of debt on prospective students and their participation,we need to go beyond the question of whether or not they decide to enter higher education.Such a focus masks other more subtle and hidden disadvantages encountered by prospectivestudents from low-income backgrounds. This is because such students exhibit a verycomplex web of attitudes towards money and debt, and employ a range of strategies for debtavoidance and to reduce the costs of higher education (Forsyth and Furlong 2000, 2003).Such strategies include choosing to stay in the family home whilst at university, applying toa university in an area with cheap living costs or where there are good term-time employ-ment opportunities, or picking a shorter course or particular subject at a less prestigious

Studies in Higher Education 407

institution. All these financial coping mechanisms in turn affect which university a studentchooses.

This article investigates whether choice of university and degree is constrained byconcerns about the cost of higher education, and whether this constraint is greater amongstudents from lower social class families compared to those from the middle and upperclasses. It also considers what attitudes drive these financially-constrained decisions, distin-guishing between those that reflect different valuations of the returns of higher educationand those based on differences in attitudes towards debt. If an unwillingness to enter into alarge amount of debt drives these decisions – and following Callender and Jackson (2005),if this is particularly so for students from low-income families – then the new variable feesmay exacerbate class differentials in higher education. Finally, the article assesses whetherfear of debt deters people from choosing certain types of qualification and pursuing certainsubject areas, particularly for students from low-income families.

By way of introduction, the article examines current government rhetoric on studentchoice and some of the assumptions underpinning their thinking. It then explores existingresearch on student choice, especially the role of material constraints, students’ methods fordealing with limited resources and concerns about accumulating debt. Following this, thearticle reports on the main findings from the survey, and demonstrates how financial issuesand fear of debt constrain the choice of university for would-be students from low socialclasses, but not for those from other classes (choice of qualification does not show thisdynamic, however). Finally, the article considers some implications of the findings forstudent funding policies.

Government rhetoric on student choice

The White Paper, ‘The future of higher education’ (Department for Education and Skills,2003a) framed the issue of student choice in very narrow terms, reflecting the notion ofhigher education as a market and students as consumers. Here choice is about competitionwithin higher education, legitimising an expanding higher education quasi-market.

The Government believes that student choice will be an increasingly important driver ofteaching quality, as students choose the good-quality courses … that … will give them thehigher-level skills that they will need during their working life. But student choice can onlydrive quality up successfully if it is underpinned by robust information. (Department forEducation and Skills, 2003a, para. 4.1, 47)

Access to such information, however, raises broader issues about both the disseminationof knowledge and flows of information and, in turn, about respondents’ cultural capital andsocial networks. The above quote assumes that having information and the correct sort ofinformation is central to decision making. Yet, this assumption ignores the whole contextwithin which decisions and choices are made. As Hutchings (2003, 98) observed in relationto initial entry decisions:

… the relationship between information and decision-making appears much less straightfor-ward than is assumed … People having access to identical information about higher educationmay construct it to come to entirely different decisions about whether or not to apply to univer-sity. These reflect their perceptions of the providers of the information, as well as a whole rangeof contextual and identity factors.

In addition, this assumption fails to recognise that some sources of information areconsidered more trustworthy than others. In our study, students from lower social classes


relied more heavily on unofficial and informal sources of information such as theirfriends and family, the grapevine, word of mouth and informal networks (Callender2003). Unlike their wealthier peers, they were less likely to supplement this with knowl-edge derived from official and formal information sources such as the government orhigher education institutions. Consequently, the government’s strategy of filling the‘information gap’ primarily through formal sources of knowledge may not have thedesired effects.

The White Paper’s accompanying document, ‘Widening Participation in HigherEducation’ (Department for Education and Skills 2003b), similarly emphasises the need forstudents to ‘make well-informed choices’ and for better information, advice and guidance,as does the government’s most recent review of widening participation policy (Departmentfor Education and Skills 2006). It discusses the social class inequalities in patterns of highereducation applications, and the ‘under-representation’ of students from low-income back-grounds applying to the most prestigious universities. The problem, it concludes, is one ofaspiration: with ‘a significant number choosing not to apply to those universities for whichthere is strong competition for places, but which may provide a good match to their talents’(Department for Education and Skills 2003b, 11). The rhetoric again calls on the languageof markets but places the mismatch, or market failure, in a lack of aspiration. The solution,therefore, ‘lies in finding better ways to reach out to potential students from a widerrange of backgrounds and encourage them to apply’ (Department for Education and Skills2003b, 12).

Underpinning the first quote from ‘Widening Participation’ is the assumption that, ifstudents behaved appropriately, the number and quality of applicants will match thenumber and quality of places available: an unrealistic proposition. Not all students canapply to, or will be accepted to, the elite institutions. ‘Being realistic’ and choosing alocal institution, or one other than the most selective, may be a better choice for some,and make more sense. This assumption also characterises students as making ‘irrational’choices. This type of discourse locates their lack of aspiration, and the concomitant lackof self-esteem, as individual problems, ignoring the structural and cultural factors andinequalities which influence their aspirations. It espouses a very narrow notion of studentchoice and a very superficial understanding of how and why students ‘limit’ theirchoices.

Both these two government documents link student choice to financial issues – hencethe reform of student support. The government acknowledges that with the introduction ofvariable fees, ‘students from low-income backgrounds, and their families, will be concernedabout the affordability of studying for a degree’ (Department for Education and Skills2003b, 18). Despite the new funding system of grants and deferred fees repaid by low-interest student loans ‘… there is a risk that potential students may be concerned aboutthe level of debt they may incur and perceive that higher education is not affordable’(Department for Education and Skills 2003b, 18). It justifies its new policies by calling uponarguments about the financial returns of higher education. Yet, it ignores how opportunitiesand constraints affect investment decisions and the accrual of benefits. Moreover, the policyrhetoric focuses exclusively on higher education as a private investment for private returns,rather than as a public investment for public returns.

In short, concerns about affordability and debt are portrayed in these governmentdocuments both as financial barriers to participation and as misplaced perceptions. Oncethese are removed, via new student funding provision and better information, equalaccess and widening participation will follow. As we will show, this is not necessarilythe case.


The research evidence

Choices are bounded by the framework of opportunities and constraints the person finds herselfin, her external circumstances. However, … she is also circumscribed by an internalized frame-work which makes some possibilities inconceivable, others improbable and a limited rangeacceptable. (Reay, David, and Ball 2005, 27)

There is a growing body of research examining the complex social, economic and culturalfactors and inequalities underpinning working class educational ‘choices’, including theirchoice of higher education institution, subject and qualification. Existing studies suggestthat financial concerns play a major role in the decision-making process of where and whatto study (Connor et al. 1999, 2001; Forsyth and Furlong, 2000, 2003; Archer, Hutchings,and Ross 2003; Reay, David, and Ball 2005). Similarly, there is a consensus in this literaturethat prospective students from lower socio-economic backgrounds are more likely thanthose from better-off families to report that their choices are constrained by the costs ofhigher education. In addition, several of these studies cite fear of debt and the prospects ofbuilding up large debts, particularly student loan debt, as a factor impacting on their choices,especially for those from low socio-economic groups.

For example, Forsyth and Furlong’s longitudinal (2000, 2003) study of Scottish disad-vantaged young people found that those who decide to enter higher education, limit theiroptions of where and what to study because of the extra financial, geographical and socialbarriers they face. They are more likely than their more advantaged peers to enrol in lessadvanced and prestigious courses at lower status institutions. In particular, because of a lackof funds and debt aversion, they opt for shorter courses, nearer their home and in subjectsthat guaranteed a job at the end of their course. Knowles (2000) and Connor et al. (1999)found similar behaviour among low-income university applicants in response to the costs ofhigher education.

A common strategy used by students concerned about the costs of higher education anddebt is to live near or at the parental home while studying. This strategy restricts students’choices, particularly those of low-income students, from minority ethnic groups and inLondon (Callender and Wilkinson 2003). Reay, David, and Ball (2005) point to workingclass localism. Material circumstances meant that most working class higher educationapplicants in their study operated within a narrow circumscribed space of choice, where thelocation of the institution was crucial because of the costs of accommodation and commut-ing. ‘Geography determines choice for a majority of working class students’ (Reay, David,and Ball 2005, 86), but not for their middle class peers.

Over the last decade, the proportion of students living at home rose from 12 to 20%,while the proportion living in university accommodation fell by the same proportion from35 to 27% (Ramsden 2006). The average distance between a student’s home and their highereducation institution is correlated with their social class. Students from professional familiestravel the longest distances and those from unskilled the shortest (Higher Education FundingCouncil for England 2001) and this pattern is becoming more entrenched (Farr 2001).

Living at or near home particularly restricts the options of students from low-incomefamilies living in regions with limited higher education provision (Forsyth and Furlong2003). In the UK, there is no one body with the power or responsibility to oversee compre-hensive provision or to ensure a full range of courses are offered in all regions. A studentwho has to attend their local university will have more limited choices, and may not be ableto take the subject or course they want.

Living at home has other consequences for students once at university, and once theygraduate. For example, new research on students’ friendships (Callender and Day Slater,


forthcoming) show that students living at home found it more difficult to make friends, hadless enjoyable and diverse social lives, and were less involved in university-based activitiesthan those living away from home. Most frequently, they relied on their pre-universityfriends, rather than branching out and making new university-based friendships andnetworks. So studying became an extension of their existing lives rather than a break fromit. Consequently, many of these students missed out on certain social aspects of universitylife, and on learning through informal interaction and socialising with their fellow students.In addition, Furlong and Cartmel (2005) reveal that students who remained in their localitywhile at university were discouraged from operating in a national graduate labour market,and ended up with lower paying local jobs. Thus, we see how social mobility, a key role ofhigher education, in part, is predicated on geographical mobility.

A further means of dealing with the costs of higher education and reducing borrowingis to get a part-time job while studying. In the UK, well over a half of all students have paidwork during term-time, and a disproportionate number come from the most disadvantagedbackgrounds. As a recent study showed, they worked because they needed the money for‘basic essentials’; ‘could not manage on their student loans’; and could not ‘rely on theirfamilies for support’, unlike their more affluent peers. And a large minority worked to avoidbuilding up debt (Van Dyke, Little, and Callender 2005).

The decision to work is not an alternative to living at home, but is positively associatedwith it: those living at home have the highest term-time employment rates (Finch et al.2006; Callender and Wilkinson 2003). Some students decide to go to their local universitypartly because they want to continue in the jobs they had before entering university.

Term-time working, like living at home, is similarly restrictive. Most serious is its detri-mental effects on students’ academic attainment. The more hours students work, the greaterthe detrimental effect on their course marks and final degree results, even after controllingfor their academic ability and other factors that might affect their results. Consequently,students working the average number of hours a week (15 hours) are a third less like to geta first or upper second class degree result than an identical non-working student (Callender2008). Purcell et al.’s (2005) research on graduates confirms the impact of term-time workon academic attainment. They also show that students’ poorer degree results, arising fromterm-time employment, led to lower paid jobs on graduation which harmed their careers,especially students from the lowest social classes.

These studies provide invaluable insights into the relationship between student choiceand financial constraints. However, none of them single out and quantify the links betweenstudents’ attitudes towards debt and how these, and their debt avoidance strategies, impacton the choices of students from different social classes. Nor do they incorporate prospectivestudents’ wide- ranging attitudes to debt. There are no comprehensive studies in the Britainthat do this in a methodologically robust way, controlling for demographics, orientationtowards the benefits of going to university, encouragement received from family andfriends, and other factors. Ideally, such a study would be longitudinal, tracking individualsfrom the whole of the UK and of all social classes over time.

The absence of suitable data sets to conduct such studies means that it is not possible toreach any firm conclusions about the impact of debt on prospective students’ actual behav-iour, choices and decision-making. However, it is possible to explore potential students’attitudes toward debt from cross-sectional studies. Yet, it is acknowledged that the relation-ship between attitudes and actual behaviour is not clear, given both cognitive dissonancetheory (Festinger 1962) and self-perception theory (Bem 1972).

The gaps in the existing research prompted this study. It is not the ideal, longitudinalstudy. However, unlike other research, it focuses on prospective students, and specifically


explores the relationship between their attitudes to debt and how their decisions aboutuniversity are constrained by financial issues – indeed it is the first study to do so. This isparticularly important, given the new student funding system’s increasing reliance onstudent loans, and the likely increase in student debt with the introduction of variablefees.

The study

The survey

In 2002, questionnaires were administered to final-year students from a stratified randomsample of British further education colleges and sixth forms, all of whom were studying forqualifications that permit entry to university. All schools (41) and (60) colleges at the 101sampling points agreed to take part; 81% of the schools and colleges returned a batch ofcompleted questionnaires. Teachers distributed an in-class questionnaire; there was a 55%response rate among the pupils (n = 1954). Final data were weighted to the national profileof educational establishment and qualification type. More details including the samplingstrategy may be found in Callender (2003).

The sample

Table 1 provides a summary of the sample characteristics of those students who had decidedto apply to university. The majority of the sample fell into the following categories: aged 21or under (90%); white (77%); and single and childless (93%). Women outnumbered men(60% compared to 40%). Over half the sample (59%) came from families where the mainbreadwinner was either ‘managerial and professional’, ‘intermediate’ or ‘small employer’;just over a quarter (26%) were from the other three social classes; data on class were notavailable for 15% of the sample.

Two-thirds of the respondents were studying in the further education sector, either attend-ing a further education or sixth form college – slightly above the UK average (Table 1). Justover a quarter of the sample attended state secondary schools, leaving less than one in tenin an independent school. More than half of all respondents were taking A- or A/S-levels orScottish Highers – the most common academic Level 3 qualifications – with most of the reststudying for some type of Level 3 vocational qualification. Just under one-quarter of thosestudying A-levels were predicted to achieve relatively high grades (BBC+).

Table 1 also shows the data broken down by low, medium and high social class (usingthe National Statistics Socio-economic Classification). Around one-third (33%) of thosefrom the managerial and professional class were expecting BBC+ grades or higher in A/AS-levels/Scottish Highers, compared with around one-in-seven (15%) of those from low-income families where the main earner in the household was in a semi-routine or routineprofession, or had never worked, or was in long-term unemployment. Equally, around one-in-seven (15%) of those in the higher social class attended an independent school, comparedto one-in-fifty (2%) of those from low-income families.

Measuring key variables

Choice of university

Respondents were asked the following: ‘To what extent has the cost of going to universityaffected any of your decisions or ideas about the following? Because of the cost I am going


Table 1. Sample characteristics of students who had decided to go to university.

Social class1

Characteristic Low SES (%) Middle SES (%) High SES (%) Total (%)

GenderMale 33 36 45 40Female 67 64 55 60

AgeUnder 21 16 9 6 90Over 21 84 91 94 10

Ethnic originWhite 68 77 88 77Non-white 32 23 12 23

Social classManagerial and professional 100 39Intermediate 100 6Small employee 100 14Lower supervisory and technical 100 8Semi-routine and routine 100 13Never worked/long-term unemployed 100 5Missing 15

Marital statusSingle 91 92 95 93Married/co-habiting 6 5 4 5Widowed/separated/divorced 3 1 1 1Not stated 0 1 0 1

Family typeSingle, childless 87 94 95 93Couple, childless 6 1 1 2Single, living with children 1 1 1 1Couple, living with children 6 4 3 3

Type of educational institutionState secondary high school 26 27 33 29Independent high school 2 8 15 9Sixth form colleges 10 8 12 10Further Education college2 63 57 40 52

Qualification aimA/AS-levels/Scottish Highers3 38 56 67 56NVQ/GNVQ/SVQ Level 3/AVCEs4 52 35 27 37Access course 6 4 3 4Other FE qualification 4 4 3 3

Expected A-Level/Highers Grade5

=>280 (BBC+) 14 21 33 231–279 (<BBC) 14 21 21 19


to or I am thinking about: …’ Respondents could tick any of the following statements thatapplied to them:

● Applying to universities nearer my home.● Taking a subject with better employment prospects.● Living at home with my parents while at university.● Applying to universities in areas where there are good opportunities for term-time

employment.● Applying to universities in areas where the cost of living is lower.● Taking a shorter course.● Doing a vocational job-related course rather than an academic course.● Applying to a ‘new’ university rather than an ‘old’ university.● Doing a part-time course.

Choice of qualification and subject

Respondents who had decided to apply to university were first asked what type of qualifi-cation they hoped to pursue (first degree, foundation degree, HND/HNC, Dip HE, other)and what subject they hoped to take (a wide range of options were given).

Measuring perceived benefits of going to university

Student attitudes towards the social, lifestyle and experiential benefits of going to universitywere measured using three Likert statements [strongly agree, agree, neither agree nordisagree, disagree, strongly disagree]:

Table 1. (Continued.)

Social class1

Characteristic Low SES (%) Middle SES (%) High SES (%) Total (%)

Not stated 8 11 11 11Not doing A-Level/Highers 64 46 36 47

N (weighted) 231 379 529 1348

Notes: The total includes all respondents who answered these questions; the class breakdown omits those for whom no data on social class were available.1The measure of social class measure was derived from a variant of the UK’s Office of National Statistics’ Social Economic Class schema. The ‘lower-income group’ was from a family where the main earner in the household was in a semi-routine or routine profession, or had never worked, or was in long-term unemployment. The ‘medium class group’ was comprised of those where the chief earner was in an intermediate or lower-supervisory/technical occupation or was a small employer or own account worker. The ‘upper class group’ contained those where the main earner was in managerial or professional employment. However, one should not treat these categories as strictly hierarchically ordered, nor reflecting mutual exclusivity in terms of bands of income or other criteria; there may be significant overlap on many of the criteria that determine social class between individuals in each of the groups2Further Education Colleges (FEC) are major providers of post-compulsory academic and vocational education serving their local community.3A-Levels are a national General Certificate of Education academic qualification usually taken in the optional final two years of high schooling and are traditionally, a prerequisite for university entry. Scottish Highers are the equivalent in Scotland.4National Vocational Qualifications are vocational awards in England and Wales that are achieved through assessment and training. In Scotland they are known as Scottish Vocational Qualification (SVQ).5The scores are based on the grades achieved in A-Levels or equivalent qualifications


● Going to university is a worthwhile experience.● One of the best aspects of going to university is developing yourself as a person.● Some of the best aspects of going to university are meeting new people and the social life.

Attitudes towards long-term career and financial benefits were measured using twostatements:

● You need a university degree to get a decent job.● In the long term you benefit financially from going to university.

An index was created for each attitude by adding up respondents’ responses to each ofthe relevant items, thus producing one indicator for attitudes towards social benefits andanother indicator for attitudes towards financial benefits.

Measuring debt attitudes

Existing studies adopt simplistic measures of debt, such as whether ‘worries about debt’affected a particular decision. Such measures do not capture in any detail how students feelabout debt and what it means to them. In our measure of debt, we try to tap these beliefsusing validated indicators. Two aspects of debt are measured: general levels of fear of debt,and a cost–benefit balance judgement concerning university. General fear of debt wasmeasured using the following attitude statements:

● Owing money is basically wrong.● There is no excuse for borrowing money.● You should always save up first before buying something.

Responses were combined into a single index using exploratory factor analysis, saving theresulting factor scores into one variable (see Callender and Jackson 2005).

The same procedure generated a second variable that reflected a balance of the perceivedcosts and perceived benefits of going to university. The attitude statements used were:

● Borrowing money to pay for a university education is a good investment.● Student loans are a good thing because it allows students to enjoy university life.● Students do not worry about their debts while at university because they will get

well-paid jobs when they graduate.● It is not worth getting in debt just so you can get a degree (recoded).

Together these measures solicit some kind of balance of their perceptions of the debts theymight accrue against their attitudes towards the short-term and longer-term benefits ofhigher education.

Results

Descriptive statistics

The influence of higher education costs on choice of university

Table 2 shows that respondents reported that financial issues influenced some decisions farmore than others. A half of all respondents were considering a university nearer their home


for this reason, but only 2% were thinking of doing a part-time course rather than a full timeone. Family was a source of financial support for a good proportion of respondents, withanother third considering living at home while at university. In addition, just under a thirdof entrants had decided to apply or was thinking of applying to universities in areas wherethere were good opportunities for term-time employment (29%). Finances also influencedthe subject they were intending to study: over a third were planning to take a subject withgood employment prospects (37%); while smaller proportions reported thinking abouttaking a shorter course (6%) and doing a vocational job-related course rather than anacademic course (7%).

Class was an important factor in these decisions (Table 2). Two-in-three of those fromlower-income families reported that they were considering applying to a university nearertheir family home to save money; this compared with two-in-five of those from the highersocial class.

Choice of qualification by social class

Table 3 shows that most respondents who had decided to apply to university hoped topursue a first degree. This is particularly the case for students from the medium and highersocial classes. Issues about subject choice are discussed later in the article, in the concludingsection.

Modelling choice of university

So far we have seen that significant numbers of students intending to go to university statedthat cost had influenced their decisions about which university to apply to and what to study.We have also seen that students from lower-income families were more likely to report afinancial constraint on university choice.

Table 2. The cost of university and key higher education decisions, by social class.

To what extent has the cost of going to university affected any of your decisions or ideas about the following?

Percentage who said cost had made them consider each

decision

Because of the cost I am going to, or I am thinking about:Low SES

Middle SES

High SES Total

Applying to universities nearer my home** 66 56 39 50Taking a subject with better employment prospects 43 41 37 37Living at home with my parents while at university** 43 35 24 32Applying to universities in areas where there are good opportunities

for term-time employment35 30 31 29

Applying to universities in areas where the cost of living is lower 28 23 27 26Taking a shorter course 8 6 4 6Doing a vocational job-related course rather than an academic course 5 7 8 7Applying to a ‘new’ university rather than an ‘old’ university 3 7 8 7Doing a part-time course 3 2 1 2

Notes: Base n = 1348.The total includes all respondents who answered these questions; the class breakdown omits those for whom nodata on social class were available**Statistically significant at the 0.01 level according to the chi-square test of independence.


In the rest of this section we investigate the factors associated students’ choices. Wefocus on social class, attitudes towards costs and benefits of higher education, andgeneral fear of debt. Do perceptions of the cost of higher education drive the desire toreduce the amount of debt one accrues by adopting cost saving measures? Or is the issueone of perceived benefit? And are perceptions of costs and benefits important inconstraining university choice in each of the social class groups? We examine whetherdebt aversion plays a particularly strong role for the lowest social classes relative to theother classes.

Social class and attitudes towards costs, benefits and debt

Before examining whether fear of debt and other factors were associated with decisions onhigher education, it was important to establish whether fear of debt, the perception of bene-fits and the balance of benefit against cost were each related to social class.

Table 4 shows that those from the lower-income group were slightly more fearful of debtthan those in the middle and upper classes. Equally, they were more likely to rate the costsof going to university higher than the benefits. However, there were no statistically signifi-cant differences in attitudes towards either the social benefits of higher education or thelong-term financial benefits.

The strategy for analysis

We estimated logistic regression models to predict whether respondents had or had notmade each of a number of different decisions because of the cost of going to university.We tested two models. The first specified main effects of: (a) general fear of debt, (b)

Table 3. Choice of qualification and subject by social class.

Low SES (%) Middle SES (%) High SES (%) Total (%)

First degree (e.g. BA, BSc)** 64 81 85 79Foundation degree 6 3 1 2HND/HNC** 24 11 8 12Dip HE 5 5 3 4Other 2 1 3 2Medicine – pure 7 7 9 8Science 3 5 6 5Maths, engineering and technology 16 14 11 13Business 19 18 17 18Social sciences 13 13 16 14Education 15 9 9 10Law 3 5 4 4Humanities** 3 9 11 9Creative arts 6 6 5 6Other 8 8 10 9Medicine – nursing** 7 7 4 5

Notes: Base n = 1348.HND, Higher National Diploma; HNC, Higher National Certificate; DipHE, Diploma in Higher Education.**Statistically significant at the 0.01 level according to the chi-square test of independence.


cost-benefit balance, (c) social benefits and (d) financial benefits. If perceived benefitswere statistically significant predictors of the decisions, then those who felt the social andfinancial benefits to be relatively low tended to pick a university and/or course thatentailed entering into less debt. If the cost-benefit balance was important, then the issuewas more of balancing up the pros and cons. Finally, if general debt aversion was a statis-tically significant predictor then the issue was more one of sensitivity to the cost of highereducation.

The second model specified interaction effects between each of these four attitudes andsocial class. It was hypothesised that the effects of: (a) general fear of debt, (b) cost-benefitbalance, (c) social benefits and (d) financial benefits might only obtain for those fromlower-income families. For example, because this group has fewer resources, they might besensitive to the cost of going to university.

We included the following explanatory variables into each logistic regression model:social class, expecting financial support while at university from the family, educationalinstitution, gender, age, ethnicity, whether the mother had been to university, educationalachievement, degree of encouragement to go received from family and friends, and havinga good idea of what university is like.

Table 4. Associations between social class and (a) general debt aversion, (b) cost and benefitbalance of university, (c) social benefits of university, and (d) long-term financial benefits ofuniversity.

95% CI for mean

Social class Mean SD SE

Lower bound

Upper bound Minimum Maximum

General debt aversion [high scores = relaxed about debt] F = 9.478; p = <0.0005

LowMediumHigh

−0.0960.0940.143

0.8290.7670.782

0.0480.0330.031

−0.1910.0290.081

−0.0010.1600.205

−2.192−2.192−2.192

1.8171.8171.817

Balance between the costs and benefits of going to university [high scores = cost outweighs the benefit] F = 5.738;p = 0.003

LowMediumHigh

0.132−0.047−0.044

0.8600.7700.804

0.0500.0340.032

0.034−0.115−0.108

0.2310.0200.019

−2.126−1.955−2.126

2.2782.1252.278

Attitudes towards the social, lifestyle and experiential benefits of university [high scores = low benefits] F = 1.646; p = 0.193

LowMediumHigh

5.4405.4735.312

1.5791.6191.595

0.0890.0690.063

5.2665.3365.189

5.6155.6095.435

333

121314

Attitudes towards the long-term career and financial benefits of university [high scores = low benefits] F = 1.164; p = 0.313

LowMediumHigh

4.7044.8954.796

1.6921.8801.799

0.0950.0810.071

4.5164.7374.657

4.8915.0544.936

222

101010

Notes: SD, standard deviation; SE, standard error; CI, confidence interval.


1. Applying to universities near the family home

Table 5 shows that, even after controlling for a range of factors, debt aversion (β = −0.252; p = 0.040), and attitudes towards both the social (β = 0.191; p = 0.006) and finan-cial (β = −0.112; p = 0.044) benefits of higher education, were associated with a reporteddecision to apply to a university nearer the family home (for financial reasons). Thosewho were more debt averse were more likely to state this particular decision. And thosewho were more negative towards the social, lifestyle and experiential benefits of going touniversity were more likely to constrain their choice of institution for reasons of cost.Those who were more positive about long-term financial benefits were more likely toapply to a university near their family home (although one must bear in mind the range ofthings being held constant). Other statistically significant covariates were: educationalattainment, pursuing qualifications other than A-levels/Scottish Highers/AS-levels andgoing to an independent school. So, students who attended an independent school wereless likely to indicate a financially-motivated decision compared to students from thefurther education sector. Similarly, students taking qualifications other than A-levels andAS-levels were more likely to indicate financial motivations than those not taking thesequalifications.

Looking across the sample as a whole, the decision to apply to higher education institu-tions near the family home (because of the cost) was associated with a reluctance to incurdebt, with negative perceptions of the social benefits of university, and with positiveperceptions of the financial benefits of higher education. Students who made this decisionseemed to be instrumental about higher education, and concerned about the debt they werelikely to accrue.

What about interaction effects? The only statistically significant interaction effectconcerned cost-benefit balance. The cost-benefit balance had a strong effect for students fromlower-income families (β = 0.744: Table 5), a weak effect for students from the medium socialclass (β = 0.196), and no effect for those from the upper social class (β = −0.085). This meantthat the cost–benefit balance was a factor only for those from low-income families: for thisgroup, judging the costs to outweigh the benefits was a predictor of applying to a universitynearer the family home.

2. Taking a subject with better employment prospects

This was the second most frequently reported decision, and statistical modelling showedthat attitudes towards the financial and social benefits of university were important. Ifstudents felt that the benefits of higher education were generally high, then they were lesslikely to constrain their choice of subject purely because of the cost of university. There wasno effect for general fear of debt or the cost-benefit balance, nor were there significant inter-action effects. For brevity, the parameter estimates are not presented here.

3. Living at home with parents while at university

Attitudes towards the benefits of university continued to be important in this decision, butthis time just the social benefits. The more students were positive about the experience ofgoing to university, the less likely they were to say they were going to live at home withtheir parents. Looking across the sample as a whole, fear of debt was not a statisticallysignificant predictor of living in the family home while at university for financial reasons.There were no significant interaction effects.

Studies in Higher Education 419Ta

ble

5.L

ogis

tic

regr

essi

on m

odel

ling

: ap

plyi

ng t

o a

univ

ersi

ty n

ear

the

fam

ily h

ome

(for

fina

ncia

l re

ason

s).

Mod

el I

Mod

el I

I

Var

iabl

esB

SE

pE

xp(B

)B

SE

pE

xp(B

)

Con

stan

t0.

146

0.70

50.

836

1.15

70.

658

1.24

30.

596

1.93

1S

ocia

l cl

ass

Low

[re

fere

nt]

Med

ium

0.03

40.

252

0.89

21.

035

0.04

51.

322

0.97

31.

046

Hig

h−0

.246

0.25

60.

336

0.78

2−1

.245

1.27

90.

331

0.28

8D

ebt

aver

sion

[hi

gh =

rel

axed

abo

ut d

ebt]

−0.2

52*

0.12

30.

040

0.77

7−0

.291

0.30

90.

346

0.74

8B

alan

ce o

f co

sts

and

bene

fits

[hi

gh =

cos

ts o

utw

eigh

the

ben

efit

s]0.

133

0.12

00.

265

1.14

30.

744*

0.33

60.

027

2.10

4P

erce

ptio

ns o

f so

cial

ben

efit

s [h

igh

= l

ow b

enef

its]

0.19

1†0.

069

0.00

61.

210

0.31

40.

177

0.07

61.

369

Per

cept

ions

of

fina

ncia

l be

nefi

ts [

high

= l

ow b

enef

its]

−0.1

12*

0.05

60.

044

0.89

4−0

.350

*0.

169

0.03

80.

704

Inte

ract

ion

effe

cts:

Med

ium

soc

ial

clas

s w

ith

…D

ebt

aver

sion

−0.0

610.

376

0.87

10.

941

Cos

t/be

nefi

t ba

lanc

e of

uni

vers

ity

−0.5

480.

390

0.16

00.

578

Soc

ial

bene

fits

−0.2

470.

209

0.23

70.

781

Fin

anci

al b

enef

its

0.27

70.

195

0.15

41.

320

Inte

ract

ion

effe

cts:

Hig

h so

cial

cla

ss w

ith

…D

ebt

aver

sion

0.12

60.

354

0.72

21.

134

Cos

t/be

nefi

t ba

lanc

e of

uni

vers

ity

−0.8

29*

0.37

50.

027

0.43

6S

ocia

l be

nefi

ts−0

.075

0.19

70.

703

0.92

8F

inan

cial

ben

efit

s0.

287

0.18

40.

118

1.33

3E

xpec

ting

to

rece

ive

fina

ncia

l su

ppor

t fr

om f

amil

y [r

efer

ent

= n

o]−0

.073

0.18

80.

697

0.93

0−0

.098

0.19

20.

609

0.90

6T

ype

of i

nsti

tuti

on a

tten

ded

[ref

eren

t =

fur

ther

edu

cati

on s

ecto

r]In

depe

nden

t sc

hool

−0.9

03†

0.35

30.

010

0.40

6−0

.949

†0.

358

0.00

80.

387

Sta

te s

choo

l−0

.165

0.21

90.

451

0.84

8−0

.202

0.22

40.

368

0.81

7G

ende

r [r

efer

ent

= m

ale]

0.34

00.

180

0.05

91.

405

0.32

50.

185

0.08

01.

383

Eth

nici

ty [

refe

rent

= n

on-w

hite

]0.

229

0.22

50.

308

1.25

80.

318

0.23

30.

173

1.37

4A

ge [

refe

rent

= o

ver

21]

−0.3

600.

343

0.29

50.

698

−0.3

270.

351

0.35

20.

721

420 C. Callender and J. JacksonTa

ble

5.(C

onti

nued

.)

Mod

el I

Mod

el I

I

Var

iabl

esB

SE

pE

xp(B

)B

SE

pE

xp(B

)

Mem

ber

of f

amil

y be

en to

uni

vers

ity

[ref

eren

t = m

othe

r no

t bee

n to

uni

vers

ity]

−0.4

42*

0.22

30.

047

0.64

3−0

.435

0.22

50.

054

0.64

7P

redi

cted

A-l

evel

gra

des

– [h

igh

= h

igh

grad

es;

AS

lev

el o

r ‘o

ther

’ st

uden

ts

code

d as

zer

o]−0

.174

†0.

035

<0.

0005

0.84

0−0

.168

†0.

036

<0.

0005

0.84

5

Stu

dyin

g fo

r A

S l

evel

s [r

efer

ent

= n

o]−0

.798

0.43

00.

064

0.45

0−0

.759

0.43

90.

084

0.46

8S

tudy

ing

for

‘oth

er’

qual

ific

atio

n [r

efer

ent

= n

o]−0

.691

*0.

299

0.02

10.

501

−0.7

59*

0.30

70.

013

0.46

8D

egre

e of

enc

oura

gem

ent

rece

ived

fro

m f

amil

y an

d fr

iend

s [h

igh

= r

ecei

ved

muc

h en

cour

agem

ent]

0.02

60.

045

0.56

91.

026

0.02

20.

046

0.62

21.

023

Hav

ing

a go

od i

dea

of w

hat

univ

ersi

ty i

s li

ke [

high

= h

avin

g a

bad

idea

]0.

104

0.11

90.

384

1.11

00.

069

0.12

20.

574

1.07

1

Not

es:

*Sta

tist

ical

ly s

igni

fica

nt a

t th

e 0.

05 l

evel

. †S

tati

stic

ally

sig

nifi

cant

at

the

0.01

lev

el. S

E, s

tand

ard

erro

r. N

= 8

17;

531

mis

sing

.


4. Applying to universities in areas with good opportunities for term-time employment

Again, perception of the social benefits of university was a factor (β = −0.275; p <0.0005; Model I, Table 6). Those who were more negative about the experience ofuniversity were more likely to look for a university in an area with relatively good oppor-tunities for term-time employment. Perception of the financial benefits of higher educa-tion was not a statistically significant predictor. Neither was the balancing of costs andbenefits.

However, fear of debt was a factor, but only for those from lower-income families(Model II). The more concerned this group was about debt, the more likely they were toreport this particular decision.

5. Applying to universities in areas where the cost of living is lower

The more positive people (from all classes) were about the social benefits of higher education,the less likely they were to constrain their potential choice of institution (Table 7).

Again, fear of debt was a factor, but only for those from the lower social class. The moredebt averse they were, the more likely they were to make this decision.

The other decisions

The same models were estimated for the following:

● Applying to a ‘new’ university rather than an ‘old’ university.● Doing a vocational job-related course rather than an academic course.● Taking a short course.

None of the key variables was associated with these decisions.

Modelling choice of qualification and subject

The second and final section of the study considered the choice of qualification – whetherrespondents were applying to take a first degree or some other qualification (e.g. foundation,HND/HNC and Dip HE), and which subject they hoped to take (e.g. medicine, science,maths/engineering/technology, business and law). We were particularly interested in theimpact of fear of debt, and whether debt aversion has a different impact depending on thesocial class of the student.

The modelling strategy was as above. On the type of qualification, debt aversion was notan important predictor for any of the social class groups. Equally, we also found that fear ofdebt was not a factor in the choice of nearly all of the degree subjects – the only exceptionwas maths/engineering/technology, where debt-averse students were more likely to makethis choice than other students, controlling for a range of other factors.

Discussion and conclusions

Many of the potential students in our study reported that their choice of university wasconstrained by the cost of going to university. However, the constraint was more frequentamong students from lower class families than those from middle- and upper-class back-grounds. Looking across all social classes, constraint was often associated with student


Tabl

e 6.

Log

isti

c re

gres

sion

mod

elli

ng:

appl

ying

to

univ

ersi

ties

in

area

s w

here

the

re a

re g

ood

oppo

rtun

itie

s fo

r te

rm-t

ime

empl

oym

ent

(for

fina

ncia

lre

ason

s.

Mod

el I

Mod

el I

I

Var

iabl

esB

SE

pE

xp(B

)B

SE

pE

xp(B

)

Con

stan

t−0

.541

0.70

40.

442

0.58

2−0

.586

1.14

70.

609

0.55

6S

ocia

l cl

ass

Low

[re

fere

nt]

Med

ium

−0.5

46*

0.24

30.

025

0.58

0−1

.934

1.27

00.

128

0.14

5H

igh

−0.4

490.

248

0.07

00.

638

−0.4

881.

212

0.68

70.

614

Deb

t av

ersi

on [

high

= r

elax

ed a

bout

deb

t]−0

.094

0.12

10.

437

0.91

0−1

.060

*0.

309

0.00

10.

346

Bal

ance

of

cost

s an

d be

nefi

ts [

high

= c

osts

out

wei

gh th

e be

nefi

ts]

0.21

00.

114

0.06

61.

234

0.23

50.

264

0.37

41.

265

Per

cept

ions

of

soci

al b

enef

its

[hig

h =

low

ben

efit

s]−0

.275

†0.

069

<0.

0005

0.76

0−0

.337

0.17

30.

051

0.71

4P

erce

ptio

ns o

f fi

nanc

ial

bene

fits

[hi

gh =

low

ben

efit

s]0.

019

0.05

50.

722

1.02

00.

011

0.15

60.

946

1.01

1In

tera

ctio

n ef

fect

s: M

ediu

m s

ocia

l cl

ass

wit

h …

Deb

t av

ersi

on1.

048†

0.37

20.

005

2.85

2C

ost/

bene

fit

bala

nce

of u

nive

rsit

y−0

.400

0.32

80.

223

0.67

0S

ocia

l be

nefi

ts0.

213

0.20

80.

306

1.23

8F

inan

cial

ben

efit

s0.

063

0.18

40.

732

1.06

5In

tera

ctio

n ef

fect

s: H

igh

soci

al c

lass

wit

h …

Deb

t av

ersi

on1.

300†

0.35

4<

0.00

053.

669

Cos

t/be

nefi

t ba

lanc

e of

uni

vers

ity

0.18

70.

310

0.54

71.

205

Soc

ial

bene

fits

0.00

80.

196

0.96

91.

008

Fin

anci

al b

enef

its

−0.0

090.

172

0.96

00.

991

Exp

ecti

ng to

rec

eive

fin

anci

al s

uppo

rt f

rom

fam

ily

[ref

eren

t = n

o]−0

.086

0.18

40.

641

0.91

8−0

.154

0.19

10.

420

0.85

7T

ype

of i

nsti

tuti

on a

tten

ded

[ref

eren

t =

fur

ther

edu

cati

on s

ecto

r]In

depe

nden

t sc

hool

−0.3

510.

322

0.27

50.

704

−0.4

530.

329

0.16

90.

636

Sta

te s

choo

l−0

.226

0.22

10.

308

0.79

8−0

.335

0.22

70.

140

0.71

5G

ende

r [r

efer

ent

= m

ale]

−0.1

660.

176

0.34

70.

847

−0.1

590.

183

0.38

40.

853

Eth

nici

ty [

refe

rent

= n

on-w

hite

]0.

289

0.22

30.

195

1.33

50.

399

0.23

20.

086

1.49

0


Tabl

e 6.

(Con

tinu

ed.)

Mod

el I

Mod

el I

I

Var

iabl

esB

SE

pE

xp(B

)B

SE

pE

xp(B

)

Age

[re

fere

nt =

ove

r 21

]0.

752*

0.34

30.

029

2.12

00.

958†

0.36

30.

008

2.60

6M

embe

r of

fam

ily

been

to u

nive

rsit

y [r

efer

ent =

mot

her

not b

een

to u

nive

rsit

y]−0

.005

0.21

50.

982

0.99

5−0

.059

0.21

80.

786

0.94

2

Pre

dict

ed A

-lev

el g

rade

s –

[hig

h =

hig

h gr

ades

; AS

leve

l or ‘

othe

r’

stud

ents

cod

ed a

s ze

ro]

−0.0

120.

035

0.73

20.

988

−0.0

050.

036

0.88

00.

995

Stu

dyin

g fo

r A

S l

evel

s [r

efer

ent

= n

o]0.

899*

0.42

10.

033

2.45

81.

072*

0.43

40.

014

2.92

1S

tudy

ing

for

‘oth

er’

qual

ific

atio

n [r

efer

ent

= n

o]−0

.405

0.31

00.

192

0.66

7−0

.518

0.32

20.

108

0.59

6D

egre

e of

enc

oura

gem

ent r

ecei

ved

from

fam

ily

and

frie

nds

[hig

h =

rec

eive

d m

uch

enco

urag

emen

t]0.

119†

0.04

50.

008

1.12

70.

137†

0.04

60.

003

1.14

6

Hav

ing

a go

od id

ea o

f w

hat u

nive

rsit

y is

like

[hi

gh =

hav

ing

a ba

d id

ea]

0.05

70.

117

0.62

71.

058

0.07

90.

121

0.51

11.

082

Not

es:

*Sta

tist

ical

ly s

igni

fica

nt a

t th

e 0.

05 l

evel

. †S

tati

stic

ally

sig

nifi

cant

at

the

0.01

lev

el. S

E, s

tand

ard

erro

r. N

= 8

17;

531

mis

sing

.

424 C. Callender and J. JacksonTa

ble

7.L

ogis

tic

regr

essi

on m

odel

ling

: ap

plyi

ng t

o un

iver

siti

es i

n ar

eas

whe

re t

he c

ost

of l

ivin

g is

low

er.

Mod

el I

Mod

el I

I

Var

iabl

esB

SE

pE

xp(B

)B

SE

pE

xp(B

)

Con

stan

t−0

.297

0.76

00.

696

0.74

3−2

.883

*1.

375

0.03

60.

056

Soc

ial

clas

sL

ow [

refe

rent

]M

ediu

m−0

.398

0.25

90.

124

0.67

22.

536

1.42

90.

076

12.6

26H

igh

−0.2

590.

265

0.32

90.

772

1.80

91.

359

0.18

36.

105

Deb

t av

ersi

on [

high

= r

elax

ed a

bout

deb

t]−0

.230

0.13

00.

076

0.79

4−1

.852

†0.

391

<0.

0005

0.15

7B

alan

ce o

f co

sts

and

bene

fits

[hi

gh =

cos

ts o

utw

eigh

the

bene

fits

]0.

148

0.12

20.

225

1.15

90.

068

0.28

30.

809

1.07

1P

erce

ptio

ns o

f so

cial

ben

efit

s [h

igh

= l

ow b

enef

its]

−0.2

00†

0.07

30.

006

0.81

9−0

.062

0.19

10.

744

0.94

0P

erce

ptio

ns o

f fi

nanc

ial

bene

fits

[hi

gh =

low

ben

efit

s]−0

.117

*0.

060

0.05

00.

890

0.22

90.

173

0.18

61.

258

Inte

ract

ion

effe

cts:

Med

ium

soc

ial

clas

s w

ith

…D

ebt

aver

sion

1.77

0†0.

456

<0.

0005

5.87

3C

ost/

bene

fit

bala

nce

of u

nive

rsit

y0.

216

0.34

90.

535

1.24

1S

ocia

l be

nefi

ts−0

.370

0.20

30.

068

0.69

0F

inan

cial

ben

efit

s−0

.238

0.22

90.

299

0.78

9In

tera

ctio

n ef

fect

s: H

igh

soci

al c

lass

wit

h …

Deb

t av

ersi

on2.

020†

0.43

2<

0.00

057.

542

Cos

t/be

nefi

t ba

lanc

e of

uni

vers

ity

−0.0

230.

330

0.94

40.

977

Soc

ial

bene

fits

−0.0

870.

213

0.68

10.

916

Fin

anci

al b

enef

its

−0.3

710.

191

0.05

20.

690

Exp

ecti

ng to

rec

eive

fin

anci

al s

uppo

rt f

rom

fam

ily

[ref

eren

t = n

o]−0

.425

*0.

195

0.03

00.

654

−0.5

20*

0.20

40.

011

0.59

5T

ype

of i

nsti

tuti

on a

tten

ded

[ref

eren

t =

fur

ther

edu

cati

on s

ecto

r]In

depe

nden

t sc

hool

−0.3

270.

327

0.31

80.

721

−0.4

750.

335

0.15

60.

622

Sta

te s

choo

l−0

.485

*0.

235

0.03

90.

616

−0.5

69*

0.24

30.

019

0.56

6G

ende

r [r

efer

ent

= m

ale]

−0.1

750.

188

0.35

30.

839

−0.2

310.

197

0.24

10.

794

Eth

nici

ty [

refe

rent

= n

on-w

hite

]0.

053

0.23

00.

818

1.05

40.

168

0.24

40.

491

1.18

3A

ge [

refe

rent

= o

ver

21]

1.49

0†0.

446

0.00

14.

438

1.83

0†0.

511

<0.

0005

6.23

1

Studies in Higher Education 425Ta

ble

7.(C

onti

nued

.)

Mod

el I

Mod

el I

I

Var

iabl

esB

SE

pE

xp(B

)B

SE

pE

xp(B

)

Mem

ber

of f

amil

y be

en to

uni

vers

ity

[ref

eren

t = m

othe

r no

t bee

n to

uni

vers

ity]

0.08

00.

229

0.72

71.

083

0.03

80.

233

0.87

21.

038

Pre

dict

ed A

-lev

el g

rade

s –

[hig

h =

hig

h gr

ades

; AS

leve

l or ‘

othe

r’

stud

ents

cod

ed a

s ze

ro]

0.05

60.

037

0.13

11.

058

0.07

6*0.

038

0.04

81.

079

Stu

dyin

g fo

r A

S l

evel

s [r

efer

ent

= n

o]0.

708

0.44

30.

110

2.03

00.

999*

0.46

20.

031

2.71

5S

tudy

ing

for

‘oth

er’

qual

ific

atio

n [r

efer

ent

= n

o]0.

074

0.30

50.

807

1.07

70.

017

0.32

50.

958

1.01

7D

egre

e of

enc

oura

gem

ent r

ecei

ved

from

fam

ily

and

frie

nds

[hig

h =

rec

eive

d m

uch

enco

urag

emen

t]−0

.023

0.04

70.

627

0.97

8−0

.028

0.04

80.

560

0.97

2

Hav

ing

a go

od id

ea o

f w

hat u

nive

rsit

y is

like

[hi

gh =

hav

ing

a ba

d id

ea]

0.03

60.

123

0.77

11.

036

0.01

60.

128

0.90

31.

016

Not

es:

*Sta

tist

ical

ly s

igni

fica

nt a

t th

e 0.

05 l

evel

. †S

tati

stic

ally

sig

nifi

cant

at

the

0.01

lev

el. S

E, s

tand

ard

erro

r. N

= 8

17;

531

mis

sing

.


perceptions of the benefits of going to university: those individuals who saw relativelyfew benefits of university opted for strategies to reduce their costs. This finding suggeststhe importance of providing students with information on these benefits, as well as thegovernments’ emphasis on information on the diversity of provision and student fundingarrangements.

For students from low-income families, however, fear of debt played an important rolein the financial constraint of decision-making. In two key decisions – applying to universitieswith a low cost of living, and applying to universities with good opportunities for term-timeemployment – fear of debt was a significant predictor only for low-income students. Poorerstudents seemed, on average, to be more sensitive to the amount of debt they would accrue.This sensitivity was seen in the impact on their choice of university of the wish to keep thedebt they accrue to a minimum. The cost–benefit balance was a predictor of applying touniversities near the family home, but again only for students from low-income families.And, as we have seen from the review of the existing literature, all such decisions potentiallylimit students’ choices: choices which affect their experience of higher education and theirfuture careers.

Across the entire sample, perceptions of the benefits of higher education drove many ofthe decisions – judging benefits to be low meant students, on average, looked to lower theircosts. Fear of debt also meant students looked to reduce their costs by attending a universitynearer their parental home. However, for students with fewer financial resources, fear ofdebt drove two key decisions, suggesting that this group had a greater sensitivity to the costsof university than the middle- and upper-classes. Increasing the cost of higher education andlevels of debt, therefore, risks constraining university choice particularly among poorerstudents, as these students were especially sensitive to the cost of going to university andthus are more likely to view higher education as involving a debt rather than an investment.

To our surprise, however, debt aversion was not related to choice of qualification, chal-lenging the findings of other research, probably reflecting a different methodologicalapproaches and analytical techniques. One might expect that increasing higher educationcosts would steer students from low-income backgrounds away from less vocationallyoriented courses, towards courses which lead to more long-term financial security such aslaw, business studies, computing, etc. So by shifting more of the financial responsibility ofhigher education on to students and away from the state, the criteria by which students selectcourses would be transformed – away from ‘intrinsic’ goods to ‘extrinsic’ goods. However,where issues of choice and perceptions of risk interact, courses which are longer (such asmedicine) and which require a greater expenditure on books and equipment needed for suchcourses on the part of the students may be seen as unattractive. The rational (maximisingreturn) approach would be to opt for courses with minimum investment and maximumreturn, namely the shortest and cheapest courses. But we found no evidence that studentsfrom low-income families were opting for particular qualification-types because of fear ofdebt. Once the changes in student funding arrangements are implemented, however, suchbehaviour may well occur.

These findings show how potential students were much more willing to respond to fearof debt by living near their family home and pursuing a course in the subject they wanted,rather than to change the subject or the type of course they wanted to pursue. Thissuggests something about relations within families, and students’ recognition that theirfamily can support them both in kind and in cash as their costs of study rise. Similarly,low-income students do not appear to compromise their subject or course because of fearof debt, but instead seek term-time work and look for universities where the costs of livingare low.


Some changes in the funding regime incorporated in the 2004 Higher Education Act areparticularly beneficial to students from low-income families. The new grants and smallerloans for low-income student should help to reduce the current unequal distribution ofstudent loan debt, whereby the poorest students are 43% more in debt on graduation thanthe richest students. However, to what extent might the new reforms address the strategieshighlighted in this article that prospective students report they will adopt to reduce theircosts and debt? In other words, are there any provisions that might actually encourage themto apply to universities where the costs of living are higher, or where term-time employmentopportunities are not so good?

The student funding regime partially recognises regional differences in the costs of living.Specifically, student loan rates are higher for students studying in London than those studyingelsewhere, but only if they live away from home. Under the new funding regime, the Londonstudent loan rates have been increased well above inflation, in recognition of the particularlyhigh living costs. However, the new grant for low-income students, unlike the previousstudent grant, makes no such allowances. The maximum grant is £2700 irrespective of wherein England a student studies or lives. These grants, therefore, do not compensate poorerstudents for the extra costs of studying in London or in other areas in England where livingcosts are high (however, they might act as an incentive to live at home because students livingat home receive the same amount of grant as those living away from home).

In turn, this has implications for higher education institutions located in such regions, espe-cially London. It reflects existing concerns that students from low-income families who do notalready live in London, or within commuting distance, may be priced out of studying in thecapital. It is perhaps no surprise, therefore, that twice as many undergraduates in London asthose studying elsewhere live with their parents. It suggests the possibility of a London studentpopulation increasingly polarised along class, income, and ethnic lines (Callender 2004).

It is hard to predict the impact of the funding changes on students’ future propensity forundertaking paid term-time employment. The new grants may reduce very low-incomestudents’ need to work, or to work long hours. But there is little evidence for this fromresearch exploring the effects of grants on term-time working (Emmerson et al. 2005). Itseems likely, therefore, that term-time work will remain part of the higher education land-scape. Consequently, it is likely that debt averse students from low-income families willcontinue to opt for universities where term-time opportunities are good. In turn, their jobsare likely to have detrimental affects on their academic attainment, and undermine theirlonger-term career prospects.

In the run-up to the 2004 Higher Education Act, Goldthorpe (2003, 11) remarked: ‘what-ever scheme is eventually introduced, close monitoring of its effects, intended and unin-tended, should provide results of major theoretical interest’. The choices students makereflect their material constraints as well as their cultural and social capital, social percep-tions and distinctions, and forms of self-exclusion – all of which are class bound. Our find-ings show how, for low-income students, the costs of higher education were often seen as adebt rather than an investment. They also illustrate how inequalities in patterns of participa-tion are perpetuated through material constraints, especially fear of debt. This is particularlysignificant because of the new student funding system’s increasing reliance on student loandebt, and the imminent rise in student debt arising from variable tuition fees.

ReferencesArcher, L., M. Hutchings, and A. Ross, eds. 2003. Higher education and social class: Issues of

exclusion and inclusion. London: Routledge Falmer.


Bem, D.J. 1972. Self perception theory. In Advances in experimental social psychology, volume 6,ed. L. Berkowitz. New York: Academic Press.

Blanden J., and E. Fitzsimons. 2007. Expansion, efficiency and social inclusion: Are New Labour’saims compatible? In A different future: The new shape of university education in England, ed.M. Gokulsing. Lampeter: Edwin Mellen Press.

Callender, C. 2003. Attitudes to debt: School leavers’ and further education students’ attitudes todebt and their impact on participation in higher education. London: Universities UK.

Callender, C. 2004. The changing finances of students studying in London: Evidence from the 2002/03 Student Income and Expenditure Survey. London: Greater London Authority.

Callender, C. 2008. The impact of term-time employment on high education students’ academicattainment and achievement. Journal of Education Policy 23, no. 4: 359–77.

Callender, C., and S. Day Slater. Forthcoming. With a little help from my friends? Social class andsocial capital in non-traditional students’ experiences of higher education.

Callender, C., and J. Jackson. 2005. Does fear of debt deter students from higher education? Journalof Social Policy 34, no. 4: 509–40.

Callender, C., and D. Wilkinson. 2003. 2002/03 student income and expenditure survey: Students’income, expenditure and debt and changes since 1998. Research report no. 487. Nottingham:Department for Education and Skills.

Connor, H., R. Burton, R. Pearson, E. Pollard, and J. Regan. 1999. Making the right choice: Howstudents choose universities and colleges. London: CVCP.

Connor, H., S. Dawson, C. Tyers, J. Eccles, J. Regan, and J. Aston. 2001. Social class and highereducation: Issues affecting decisions on participation by lower social class groups. Researchreport RR 267. London: Department for Education and Employment.

Department for Education and Skills. 2003a. The future of higher education (White Paper). London:HMSO (Cm 5735).

Department for Education and Skills. 2003b. Widening participation in higher education. London:Department for Education and Skills.

Department for Education and Skills. 2003c. Student loans and the question of debt. London:Department for Education and Skills. http://www.dfes.gov.uk/hegateway/hereform/index.cfm.

Department for Education and Skills. 2006. Widening participation in higher education. Nottingham:Department for Education and Skills.

Emmerson, C., C. Frayne, S. McNally, and O. Silvax. 2005. Evaluation of Aimhigher: Excellencechallenge economic evaluation of opportunity bursaries. DfES Research Report RR647.Nottingham: Department for Education and Skills.

Farr, M. 2001. Home or away? A study of distance travelled to higher education 1994–1999. Widen-ing Participation and Lifelong Learning 3.1: 17–25.

Festinger, L. 1962. A theory of cognitive dissonance. London: Tavistock.Finch, S., A. Jones, J. Parfrement, A. Cebulla, H. Connor, J. Hillage, E. Pollard, C. Tyers, W. Hunt,

and G. Loukas. 2006. Student income and expenditure survey 2004/05. Research report RR725.Nottingham: Department for Education and Skills.

Forsyth, A., and A. Furlong. 2000. Socio-economic disadvantage and access to higher education.Bristol: Policy Press/Joseph Rowntree Foundation.

Forsyth, A., and A. Furlong. 2003. Losing out? Socioeconomic disadvantage and experience infurther and higher education. Bristol: Policy Press/Joseph Rowntree Foundation.

Furlong, A., and F. Cartmel. 2005. Graduates from disadvantaged families: Early labour marketexperiences. Bristol: Policy Press/Joseph Rowntree Foundation.

Goldthorpe, J. 2003. Sociology as social science and Cameral sociology: Some further thoughts.Sociology Working Papers, 2003–07, University of Oxford.

Higher Education Funding Council for England. 2001. Supply and demand in higher education.Bristol: Higher Education Funding Council for England.

Hutchings, M. 2003. Information, advice and cultural discourses of higher education. In Highereducation and social class: Issues of exclusion and inclusion, ed. L. Archer, M. Hutchings andA. Ross, 97–118. London: Routledge Falmer.

Knowles, J. 2000. Access for few? Student funding and its impact on aspirations to enter highereducation. Widening Participation and Lifelong Learning 2.1: 14–23.

Purcell, K., P. Elias, R. Davies, and N. Wilton. 2005. The class of ‘99 – A study of the early labourmarket experiences of recent graduates. Research report no. RR 691. Nottingham: Departmentfor Education and Skills.


Reay, D., M. David, and S.J. Ball. 2005. Degrees of choice: Social class, race and gender in highereducation. Stoke on Trent: Trentham Books.

Ramsden, B. 2006. Patterns of higher education institutions in the UK – Sixth report. London:Universities UK.

Van Dyke, R., B. Little, and C. Callender. 2005. Survey of higher education students’ attitudes todebt and term–time working and their impact on attainment. Bristol: Higher Education FundingCouncil for England, 169 pp. http://www.hefce.ac.uk/Pubs/rdreports/2005/rd15_05/rd15_05.pdf.

Contra in Sts

Documents

accumulation of student

accumulation of debt

average student loan

student finances

variable tuition fees

higher educationdoi

higher education0307

atsupporting students