Student Attitudes to Student Debt

JOURNAL OF

ELSEVIER

Journal of Economic Psychology 16 (1995) 663-679

S t u d e n t attitudes to s t u d e n t debt " Emma Davies, Stephen E.G. Lea

*

University of Exeter, Department of Psychology, Washington Singer Laboratories, Exeter EX4 4QG, UK Received 28 July 1992; accepted 18 February 1995

Abstract Levels of debt and attitudes towards debt were investigated in a sample of undergraduate students. Students were found to be a relatively low-income, high-debt group with relatively tolerant attitudes towards debt. Some of the variables that have been found to be associated with debt in general public samples were also found to have significant effects in the student group: these included religion, age, number of credit cards used, and more tolerant attitudes towards debt. In addition, men were more likely to be in debt than women. Variables correlated with tolerant attitudes towards debt included age, some kinds of expenditure, religion, and external locus of control. A pseudo-longitudinal design was used to examine the relationship between attitudes and debt: cohort (year of study) was taken as a proxy for time. Higher levels of debt, and greater tolerance of debt, were found in students who had been at university longer. The increase in debt occurred earlier in students' careers than the increase in tolerance towards debt. The results are interpreted in terms of a life cycle theory of economic behaviour, and a behavioural theory of attitude change. Students come from relatively prosperous socioeconomic groups but have low incomes which they perceive as temporary; to sustain their expected life style, they have to accept some level of debt. Their attitudes then adjust towards tolerance of debt so as to ensure consistency.

This paper is based on work submitted by Emma Davies in partial fulfillment of the requirements of a BA(Hons) degree in Psychology at the University of Exeter. * Corresponding author. E-mail: [email protected], Fax: +44 1392 264623, Tel.: +44 1392 264612. 0167-4870/95/$09.50 © 1995 Elsevier Science B.V. All rights reserved SSDI 0167-4870(95)00027-5

664

E. Davies, S.E.G. Lea/Journal of Economic Psychology 16 (1995) 663-679

I. Introduction

Debt is an increasing social problem in most developed countries at present. The prosperity and economic growth of the 1980s led to increased business confidence and optimism. Borrowing at all levels increased and the Governor of the Bank of England stated in 1989 that, more than ever, "credit has been available from more sources and for a greater variety of purposes" (Leigh-Pemberton, 1989). In the sustained recession of the early 1990s, many borrowers have experienced great difficulty in repaying their loans. Levels of outstanding credit in Britain rose from £65 billion in 1980 to £270 billion in 1988. Of these figures, the majority consisted of outstanding house mortgage repayments, outstanding personal consumer credit figures being £13 billion in 1980 and £46 billion in 1988 (Ford, 1988). Corresponding to the increase in problem debt has been a recent increase in research on debt by a variety of social scientists, including sociologists (e.g. Ford, 1988), lawyers (e.g. Sullivan et al., 1989), and political scientists (e.g. Berthoud and Kempson, 1992), though with surprising little contribution from economists (but see Cameron and Golby, 1992). Research on debt within economic psychology can be traced back to Katona (1975), but has expanded rapidly in recent years (e.g. Livingstone and Lunt, 1992; Lea et al., 1993; Tokunaga, 1993). Particular attention has focussed on two questions: what factors cause some people to use credit more than others, and what factors cause people to get into difficulties, so that an agreed and manageable credit arrangement becomes a debt which could not be repaid, or in the limit gives rise to a debt crisis (cf. Lea et al., 1993). Katona (1975, chapter 15) listed three reasons why an individual might display "dissaving" (expenditure exceeding income): low income, so that necessary spending cannot be met; high income combined with a high willingness to spend; and unwillingness not to spend (regardless of income). His list identifies a question that has dominated both public and academic discussion of the growing debt problem of recent years: is it solely to be explained by adverse economic circumstances, or is it partly due to psychological factors, in particular to changes in public attitudes, so that people are more tolerant of debt? Ford (1988) argues that the causes of housing debt problems are economic: "households ... default on credit commitments because they experience events that remove much or all of their income", or "household incomes remain the same but demands on income increase to a point where they cannot be met". On the other hand,


665

Livingstone and Lunt (1992) argue that attitudes to debt are strongly correlated with the extent of debt that people take on. The present study explores the interaction of debt and attitudes using a student population. Students are an interesting group to study in the context of debt for a number of reasons: 1. Students are relatively homogeneous with regard both to their incomes and to their necessary expenditures, so non-economic factors may be seen more readily than in a general public sample. 2. Students are particularly exposed to debt. Full-time study precludes full-time earning. In the UK, student finance is currently being shifted from a system of government support through grants to the majority of students, to a mixture of parental support, commercial loans, and government sponsored low-interest loans (introduced for the first time in 1989). 3. Some authors (e.g. Etzioni, 1988) have argued for a generational shift in attitudes, with traditional hostility to debt now breaking down. Current students grew up in the easy-credit 1980s and might be expected to be relatively tolerant of credit and debt. 4. From the point of view of life-cycle theories (e.g. Ando and Modigliani, 1963), students might be expected to be relatively willing to take on debt. Whatever their source of income, it is likely to be low relative to their expectations of future income: economic theories of education (e.g. Schultz, 1960) argue that studying is a form of investing in one's own human capital, and if the investment is worth while, it is presumably worth borrowing in order to make it. 5. The extent of debt that students are carrying on graduation may be an important issue in their career choices; for example, Fox (1992) shows that in the United States, undergraduate debt deters women, but not men, from entering graduate training. 6. Credit use among young adults in general is an important issue (Ford, 1990), and there is some indication that they may be high users of credit (Ford, 1992). The study had three main aims, two of which follow from the points listed above, while the third aims to make use of another feature of the student situation. First, we wanted to look at the extent and nature of debt among students, to see whether they are indeed an unusual group in their debting behaviour. Our second aim was to look at the factors associated with indebtedness in a student group, to see whether they are different from those found in general public samples (e.g. Lea et al., 1993; Toku-

666


naga, 1993). Finally, we wanted to use the student group to throw light on the relation between attitudes to debt and level of indebtedness by means of a quasi-longitudinal study. This final aim requires further explanation. Although previous studies (e.g. Lea et al., 1993; Livingstone and Lunt, 1992; Tokunaga, 1993) have reported some correlation between attitudinal variables and debting behaviour, these cannot, of course, be taken as implying that a tolerant attitude to debt causes people to get into debt. It is equally possible that causality runs in the other direction. Any one of several consistency theories of attitude change (e.g. cognitive dissonance theory, Festinger, 1962; self-perception theory, Bem, 1972) would predict that, if people cannot avoid getting into debt, they will adjust their attitudes to become more tolerant of it. Economic psychologists have been concerned with attitude-behaviour consistency in the specific field of debt. Maital (1982) argues that people who are opposed to debt nonetheless incur it, but consistency is maintained because people do not recognize some forms of credit use as involving a debt. Etzioni has suggested that "Taking out the first loan, for people who feel being in debt is a moral evil, is different from extending it or taking out a second one" (cited in Earl, 1992). For many people this occasion will occur whilst at university, since it is then that they are in charge of their own finances for the first time; furthermore loans are easy for students to get, through overdrafts, government student loans, or credit cards. The best way to investigate the direction of causality is with a longitudinal design, in which the temporal relation of changes in attitude and changes in behaviour can be observed. Longitudinal studies, however, are time-consuming and often weakened by participant attrition. The cohort structure of the student population makes it possible to carry out a pseudo-longitudinal study, in which differences between year groups are taken as a proxy for changes due to the passage of time. Informal observations suggested that final year students both carry more debt, and are more tolerant of debt, than first-year students. If attitude change causes change in indebtedness, we would expect to observe an increase of debt tolerance in younger cohorts than an increase in indebtedness; if behavioural change causes attitude change, we would expect to see indebtedness increase in younger cohorts than attitudinal tolerance. Attitude towards debt can be measured by a suitably constructed psychometric scale. However, other psychological variables may also affect debt. Livingstone and Lunt (1992) and Tokunaga (1993) found some correlation between debt and locus of control (Rotter, 1966). Walker et al. (1992) and


667

Livingstone and Lunt (1992) both report that crisis debt may be associated with a general failure to cope, and in an attempt to measure this, we also looked at a scale of stressful life events (cf. Holmes and Rahe, 1967), which Tokunaga (1993) also found to be correlated with debt. In addition, we measured a number of economic and social variables, since these will inevitably be correlated with debt and must therefore be taken into account in the analysis.

2. Method

2.1. Participants The participants consisted of 49 1st year, 40 2nd year and 51 3rd year undergraduate students at the University of Exeter; as is usual in English universities, most Exeter first degree courses last for three years. The three groups were approximately matched for gender.

2.2. Questionnaire The final questionnaire was six sides of A4 long, containing basic demographic items, three psychological scales, and questions about the respondent's financial position and financial habits. Construction of attitude scale. A list of 32 statements, 19 pertaining to debt and 13 to student debt in particular, was presented to an independent sample of 15 individuals who rated the statements according to how pro or anti they felt they were towards the subject of debt. On the basis of these ratings, 15 items were selected to form an attitude scale, 8 of which were " p r o " statements and 7 of which were "anti" statements. This scale was then tested for reliability. A further 48 independent individuals indicated the extent to which they agreed or disagreed with the statements and item analysis was performed on these results. The deletion of one " p r o " item ("owing money to parents is not really a debt") resulted in a satisfactory Cronbach's alpha of 0.79. The remaining 14 statements were included in the final questionnaire as the measure of attitude, and they are listed in Table 1. Life events scale. The Life Experiences Survey (Sarason et al., 1978) was used as a measure of stressful life events. This is an 21-item scale, adapted for use with students from the original scale described by Holmes and Rahe (1967).

668


Table 1 List of statements included on scale of attitude to debt Statement 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.

There is no excuse for borrowing money Banks should not give interest-free overdrafts to students Students have to go into debt It is O K to borrow money in order to buy food You should always save up first before buying something Debt is an integral part of today's lifestyle Students should be discouraged from using credit cards Banks should not be surprised when students incur large debts It is O K to have an overdraft if you know you can pay it off Once you are in debt it is very difficult to get out You should stay at home rather than borrow money to go out for an evening in the pub It is better to have something now and pay for it later Taking out a loan is a good thing because it allows you to enjoy life as a student Owing money is basically wrong

Locus of control scale. The original 9-item form of the Internal-External locus of control scale (Rotter, 1966) was used. Other questions. The questionnaire also included basic sociodemographic items, and items concerned with childhood economic experiences, with the extent to which subjects plan ahead, with "worry" encountered over the participants' bank balance and how they felt about going overdrawn. Finally, it included items about the amount and sources of the respondents' income, their estimated weekly expenditures on a variety of items, and their debts. Expenditure was recorded in nine categories (rent, food, clothing, entertainment, travel, books, gifts, interest on loans and other). Debts were recorded in four categories, bank overdrafts, credit card debt, debts under the government student loans system, and debts to family and friends. 2.3. Procedure The questionnaires were distributed and collected over a 2-week period in February 1992, about half way through the academic year. First year students were approached in the evenings in their halls of residence (all Exeter students have the opportunity to live in such halls during their first year and most choose to do so). The questionnaires were collected half an hour after distribution. Second year students were approached in the evenings in University self-catering fiats and also through contacts in


669

academic departments. Third year students were approached solely through contacts in academic departments. Psychology students were not included. This method ensured a 100% return rate from those who agreed to participate.

3. Results

One second year student reported a total income of £22000, over seven times the mean and over 3.5 times the next highest income (£6000). This student had a correspondingly large debt (£6100). His data have been dropped from all the analyses, to avoid distorting or masking more general trends.

3.1. Cohort differences The mean data for selected variables can be found in Table 2, broken down by year group. Income and expenditure responses have been converted to annual figures. Note that total mean expenditure does not add to total mean income plus total mean debt, because income and expenditure means are based on respondents' description of current-year flows, while debts are stock figures, based on respondents' statements of their immediate positions, and may have included debts brought forward from previous years. In addition, respondents may have been adding to or running down savings. In the entire sample 43% reported some debt. The percentages of debtors in each year were: 1st years 22.4% (n = 11), 2nd years 39% (n = 16), and 3rd years 67.3% (n = 33). Table 2 confirms the expected relationships between cohort and both tolerance of debt and levels of indebtedness: both increased with year group. The significance of these trends was tested by a one-way analysis of variance. Because socioeconomic class and locus of control were also monotically related to year group, they were included in the analysis as covariates. Both total debt and attitude showed significant differences between cohorts (F(2,120)= 7.54 and F(2,120) = 16.10 respectively, p < 0.001 in both cases); the effects of the covariates were not significant. The pattern of the relationship of debt and attitudes scores with cohort is illustrated in Fig. 1.

670


Table 2 Mean data over the entire sample and for each year group separately, on major variables of the study. All income and expenditure items have been converted to £ / y e a r Variable

Mean All years

Age (years) Class 1 = professional 7 = manual, routine Attitude score 1 = anti debt 7 = pro debt Locus of control + 21 = internal 21 = external Life events this year positive negative Income Total Expenditure Total rent food clothing entertainment travel books gifts interest other Debt Total student loans bank overdrafts credit cards family and friends

Year 1

Year 2

Year 3

19.58 1.78

18.51 1.95

19.67 1.85

20.55 1.51

4.33

3.98

4.17

4.81

2.73

4.69

2.34

1.14

2,43 1,16

3,08 1.35

2.50 1.25

1.75 0.90

-

3088

2792

3391

3140

3280 1339 386 208 927 174 117 74 9 45

2985 1684 135 195 635 124 143 55 2 12

3124 1136 515 190 827 165 198 89 4 90

3690 1160 529 236 1292 230 99 82 18 43

306 78 78 23 127

53 12 6 1 33

341 79 66 46 150

527 143 158 26 201

3.2. Patterns of debt and expenditure The four types of debt recorded were moderately inter-correlated. The highest correlation (0.33) was between government student loans and overdrafts; the lowest (0.00) between government loans and loans from family and friends. Correlations of types of debt with the totals of other types of debt varied from 0.10 (for government loans) to 0.47 (for loans


Debt (£)

Attitude

600 I [ ~ Debt

score

671

~5

Attitude 4.5

400 I

2OO

0

/ 1

3.5

-------2 Year of study

3

Fig. 1. Students' m e a n extent of debt reported, and attitudes to debt, in relation to year of study. Note that the attitude scale ran from 1 (strongly anti-debt) to 7 (strongly pro-debt).

from family and friends). As Table 2 shows, however, the development of the different kinds of debt across the years of the course was rather different. Expenditures on different categories also showed generally positive intercorrelations, except that spending on rent was negatively correlated with most other categories; this probably reflects the different patterns of spending of different year groups (apparent in Table 2), with first year students, mainly living in halls of residence, with higher rents which include some meals. The highest correlation (0.48) was between expenditures on food and travel, the lowest ( - 0 . 4 1 ) between rent and food; if spending on rent is ignored, the lowest correlation was -0.09, between books and other expenditure). Correlations between expenditure types and the totals of all other types varied from - 0 . 2 0 for rent and -0.03 for other, to 0.40 for travel.

3.3. Predictors of debt and debt attitudes Taking the sample as a whole, income and expenditure were positively (though not significantly) correlated, as were income and debt and income and attitude score (taking debt-tolerant attitudes as positive); there were significant positive correlations between debt and expenditure, debt and attitude, and expenditure and attitude. The only significant correlations involving the other two psychological scales were a negative correlation

672


between negative life events and total expenditure, and a negative correlation between external locus of control and income. The only other significant correlations with debt or attitudes were with age and year group, some of the categories of religion, worrying less about one's bank balance, and having more credit cards; all were significant for both dependent variables, though the categories of religion involved were different in the two cases. Fig. 1 suggests that level of debt and attitude to debt develop differently across the three years, with most of the change in level of debt between years 1 and 2, and most of the change in attitude between years 2 and 3. Accordingly, in the regression analyses that follow, year of study was replaced by two dummy variables, one contrasting years 2 and 3 from year 1, and the other contrasting year 3 with years 1 and 2. Stepwise logistic regression analysis was used to explore which of the other variables measured distinguished debtors from nondebtors. Four versions of the analysis were run: two used total expenditure as a regressor, and two used the nine categories of expenditure separately; in one of each of these, the attitude score was omitted. Debtors were discriminated by being older, worrying less about their bank balance, and having more credit cards. When the attitude score was included, debtors were found to have a more pro debt attitude, but there were no other changes in the variables included. Stepwise multiple regression analyses were used to explore the variables associated with level of indebtedness and attitude score. Again expenditure was included either as a total or as individual categories; for indebtedness, regressions were run with and without the attitudes measure, and for attitudes, regressions were run with and without indebtedness. For total debt, the variables retained in the equation as associated with higher levels of debt were being in years 2 or 3 of the course, worrying less about their bank balance, having more credit cards, and being atheist or agnostic (rather than protestant, the modal category) in religion; including attitude score among the regressors resulted in its being included in the model, but led to no other changes. For attitudes, the variables retained in the equation as associated with more tolerant attitudes to debt were being in year 3 of the course, being older, having a more external locus of control, and not being of a non-Christian religion. If expenditure was considered as a total, higher expenditure was associated with more tolerant attitudes to debt, and so was a higher score on the negative life experiences scale. However, when expenditure was disaggregated, higher expenditures on clothing, entertainment and other items were associated with more toler-


673

Table 3

Results of stepwise regression analysis to predict extent of i n d e b t e d n e s s a n d attitudes to debt. Entries are regression coefficients (for logistic regression in the case of the " S o m e d e b t " d e p e n d e n t v a r i a b l e ) . D e b t a n d e x p e n d i t u r e v a r i a b l e s a r e s c a l e d in £ / y e a r . T h e s y m b o l " - " i n a c o l u m n indicates that a v a r i a b l e w a s not tested in that regression; a b l a n k e n t r y indicates that the v a r i a b l e w a s tested, but not retained by the stepwise procedure Independent

Dependent v a r i a b l e

variables

Some

Attitude Total debt Y e a r of study

debt

-

1.25 * *

-

-

year 2 and 3

Level of debt -

-

1.50" *

272.9 * *

191.7

year 3

Attitude -

-

-

0.00024 *

0.41 * *

Age

0.85'**

0.65"*

Male gender

1.11"

1.22 * *

Expenditure on: food clothing entertainment other N u m b e r of credit c a r d s Religion (re: Protestant)

- 0.00039 * 0.00063 * * *

0.00064 * * *

0.00023 * *

0.00016 * *

0.00049 0.74

0.61

Atheist/Agnostic

347.3"**

307.6"**

333.0 * *

329.0 * *

Non-Christian

- 0.51

Locus of control Model c h i - s q u a r e adjusted

0.01 *

24.3%

28.2%

38.4%

F

14.29" * *

13.17" * *

10.65"* *

*p
**p<0.05;

- 0.60 * 0.01

33.50 * * * 45.79 * * *

R2

Degrees of freedom

0.51" * *

0.16 *

3

4

3,121

4,120

8,116

37.1% 13.21" * * 6,188

***p
ance of debt, but higher expenditures on food with less tolerance; the effect of negative life experiences fell well short of the criterion for inclusion. When the attitude score was included among the regressors, higher total debt was found to be significantly associated with more tolerant attitudes to debt, whether or not expenditure was disaggregated. Including attitudes led to no change in the other variables in the model when expenditure was considered as a total, but when expenditure was disaggregated, adding total debt to the model resulted in the dropping of age and of expenditure on food and other categories, though they fell little short of the criterion for inclusion. Numerical results from these regressions are given in Table 3. Since different categories of expenditure were found to have different relations

674


to attitudes, only the regressions using the separate categories of expenditure are shown.

4. Discussion

4.1. Student finance On the average, and in each year of the course except the second, students' mean expenditure exceeded their mean income. While some students may have been able to fund the difference by running down savings, few would have substantial economic resources. The rising debt shown in Fig. 1 is the inevitable result. By the third year, over two thirds of the sample were reporting some debt. This is a much higher proportion than would be found in any general population sample, and this leads to two questions: Why do students tend to get into debt more than other people, and why do some students get into debt more than others? Two straightforward answers can be offered for the first question. As Table 2 shows, the students had low mean incomes, which according to Lea et al. (1993) would lead us to expect higher levels of debt. But as Table 2 also shows, they came disproportionately from prosperous socioeconomic groups. On life-cycle grounds, and given that students are at the beginning of their careers, their social position should also predispose them to higher levels of debt, since their long-term income expectations would be high. They would also be used to a relatively comfortable lifestyle, and reluctant to give it up. To answer the second question, we turn to a discussion of the factors associated with levels of debt and with tolerant attitudes to debt.

4.2. Factors associated with debt Not all students were in debt. Debtors tended to be older, to have a more pro-debt attitude, to have several kinds of debt, to worry less about the level of their bank account, and to be male. Level of indebtedness was predicted by being in the second or third year of the course, having more credit cards, being atheist or agnostic rather than protestant in religion, and having a more pro-debt attitude. With one exception, all the variables found to be associated with debt have already been reported in studies using general public samples. The


675

exception is the effect of gender. It will not astonish anyone who has much to do with students to learn that more men than women have debts; we could plausibly relate the difference to differences either in expenditure patterns or budgeting styles, the latter explanation being more satisfactory given that expenditure patterns did not significantly predict either the presence or the extent of debt. Student samples are unusual in that the economic positions of men and women are broadly similar - it would be difficult to study gender differences in debt meaningfully in a general public sample. More surprising are the variables for which no significant association with levels of debt was found. These include income, expenditure patterns, locus of control, and life experiences, all of which have been found to be associated with levels of debt in other studies (e.g. Livingstone and Lunt, 1992; Tokunaga, 1993). As far as life events are concerned, the absence of an effect may have been due to limitations in the Sarason et al. (1978) scale (it gives equal weighting to items such as "death of a close family member" and "change of residence"), but in any case the number of stressful life events reported was low. The stress effect was predicted on the basis of results obtained by Walker et al. (1992) from a study of people living much more difficult lives in a depressed inner-city housing estate, and it is reasonable to guess that coping only collapses at levels of stress not normally experienced by students; it should also be borne in mind that these students were of relatively privileged socio-economic class (see Table 2). In the case of locus of control, it should be noted that Livingstone and Lunt (1992), who report that it is associated with levels of debt, did not retain the full scale in their analyses, but selected items. Locus of control scales have often been found to have unsatisfactory psychometric properties (Kline, 1993, p. 536). We did not pursue an item by item analysis, partly because our sample size was too small for it to be statistically safe, but mainly because locus of control did have significant associations with attitude to debt. The relationship between debt and income can be looked at in two ways. As was pointed out above, if we consider the students as a group and compare them with the general public, their behaviour was consistent with some recent studies (e.g. Lea et al., 1993) that have reported more frequent and higher debt in low-income respondents: students are a low income group and were relatively likely to be in debt. Within the student sample, however, the trend of the results was for higher incomes to be associated

676


with higher debt, consistent with data of Berthoud and Kempson (1992) and with data on dissaving discussed by Katona (1975, chapter 15). The correlation between debt and income was not significant, but this may simply reflect the relatively narrow range of student incomes; none of the variables observed were significantly correlated with income. It is notable that the one student whose data were excluded because he had an exceptionally large income also had an exceptionally large debt. There is no conflict between these different results, but they do suggest that there are two distinct processes linking income and debt, corresponding to two distinct routes into debt. On the one hand, very low income (combined with high necessary expenditure) drives people into debt, which may become chronic; this is the sort of effect reported by Lea et al. (1993), who were deliberately oversampling serious debtors. On the other hand, those with higher incomes can afford to take temporary debt, knowing they can repay it, and the higher their incomes, the higher the debt they can safely afford; this kind of effect is seen in general population samples such as those studied by Berthoud and Kempson (1992) and by Katona. Our student sample reflects both tendencies. Their low incomes predispose them to debt, but their good long term prospects (derived from high social status as well as education) mean they can regard it as temporary, and accordingly the level of debt they will tolerate is likely to increase with income, within the relatively narrow income range that students experience. This last analysis brings our results within the framework of life-cycle theory. However, questions about forward planning did not yield significant relationships with income, expenditure and debt. Most respondents claimed they did not plan ahead at all. Third years were somewhat more inclined to admit to thinking about what they would be doing after graduation, and a fair number of respondents did say they were planning for Christmas. We conclude that life-cycle considerations do not play any significant part in students' day to day financial decision making, even though we argue that they determine their overall financial strategy.

4.3. Attitudes towards debt and factors associated with them The mean levels of attitude towards debt were slightly on the side of tolerance. This result contrast sharply with those obtained by Livingstone and Lunt (1992) and Lea et al. (1993), who found general disapproval of debt. While the studies used quite different scales, the difference of results is consistent with the hypothesis that students will be more tolerant of debt

E. Davies, S.E.G. Lea/Journal of Econom& Psychology 16 (1995) 663-679

677

than the general public. This can be predicted on a number of grounds, demographic (they are relatively young), life cycle (they are in a phase of temporary low income), or attitudinal (circumstances force them into debt so they adjust attitudes to maintain consistency). The multiple regression analysis carried out on attitude score (Table 3) yielded some interesting results. It is not surprising that amounts spent on clothes and entertainment should have significant positive correlations with tolerance of debt; it is less obvious that amounts spent on food should have a negative association, especially given that expenditure on food rises sharply for the first to the second year of the course, at the same time as a large shift towards greater toleration of debt. The results suggest two distinct kinds of lifestyle, one more utilitarian and associated with dislike of debt, the other more "frivolous". This distinction is consistent with the approach to debt taken above, as a symptom of relative prosperity. External locus of control was significantly correlated with a tolerant attitude to debt, though life stress was not. Despite the difficulties over locus of control scales, therefore, we conclude that this variable deserves inclusion in future studies of the psychology of debt.

4.4. Relationship between attitudes and debt The most striking result from this study is the increase in both total debt and attitude score across the three years of study. Debt rose from an average of £53 per head in the 1st year to £527 per head in the 3rd year. In the first year, 25% of students were in debt, compared to 65% of 3rd years. Similarly, attitude score was slightly anti-debt in the 1st year but by the 3rd year it had risen to a strongly pro-debt score. However, the rises in debt and attitude scores followed different courses over the three years. Debt rose sharply from the first to the second year; attitudes changed more gradually, with the major difference coming between the second and third year. These patterns were confirmed by the regression analyses, in which the dummy variable for years 2 and 3 was consistently significant in analyses of total debt, while the dummy for year 3 was consistently significant for attitudes. Although a pseudo-longitudinal study cannot be as convincing as a truly longitudinal one, this pattern of results is exactly what is predicted from the hypothesis that, in the field of debt at least, behaviour change causes attitude rather than vice versa. Comments that students made when completing the questionnaires support this analysis: they suggested that debting behaviour changes towards the

678


end of the 1st year or the beginning of the 2nd year and that attitudes change thereafter. Several individuals reported that they remembered thinking before arriving at university that they would never go into debt, but events overrode their resolve, and now they did not think it so bad. We argue, therefore, that our results do not support a simple generational shift in attitudes to debt. Students were not at first well disposed to debt; but they found themselves in a situation where going into debt was convenient and easy, and changed their attitudes as a result. The regression analyses show that there was quite a strong association between debt and attitudes towards debt. However, apart from cohort, the other variables associated with the two were quite different (see Table 3). The variables identified as associated with tolerant attitudes to debt would have been plausible as causes of debt, but in fact had no significant association with it, and vice versa. If we accept, from the cohort effects discussed above, that a tolerant attitude to debt is partly induced by taking on debt, we also have to accept that other factors (such as expenditure patterns and locus of control) are independently associated with such attitudes.

5. Conclusion

The main features of our results are consistent with a life-cycle approach to debt. Students are in a special phase of their life cycle, and their behaviour and attitudes towards debt show some corresponding special features. The students in our sample were individuals from a prosperous background in a period of low income, which they could reasonably expect to be below their " p e r m a n e n t income" (cf. Friedman, 1957). Although they were not at first in favour of debt, to sustain an acceptable lifestyle, they found they had to go into debt. Once they had incurred debt, their attitudes to debt changed, so that they became more tolerant. It would be interesting to know what happens to these individuals when they leave university. Will they take a tolerant attitude to debt with them, and be permanently more willing to use credit than the general population? In that case, some proportion of them will probably end up as crisis debtors, as a result of unforeseen adverse circumstances. Or will they take the opportunity to pay off student debts (small in comparison with graduate incomes) and, once free of debt, revert to a more typical anti-debt attitude?


679

Acknowledgements Financial support came from the T.C. Old Bequest fund, and from Welsh Water plc through a grant to Stephen Lea and Paul Webley. References Ando, A. and F. Modigliani, 1963. The life cycle hypothesis of saving. Economic Review 53(1), part 1. Bem, D.J., 1972. 'Self perception theory'. In: L. Berkowitz (Ed.), Advances in Experimental Social Psychology. Vol 6. New York: Academic Press. Berthoud, R. and E. Kempson, 1992. Credit and Debt: The PSI Report. London: Policy Studies Institute. Cameron, S. and D. Golby, 1992. A review of policies towards household problem debt. Political Quarterly 63, 227-233. Earl, P.E., 1992. 'On the complementarity of economic applications of cognitive dissonance theory and personal construct psychology'. In: S.E.G. Lea, P. Webley, and B.M. Young (Eds.), New Directions in Economic Psychology. Aldershot: Edward Elgar. Etzioni, A., 1988. The Moral Dimension. New York: Free Press. Festinger, L., 1962. A Theory of Cognitive Dissonance. London: Tavistock. Ford, J., 1988. The Indebted Society. London: Routledge. Ford, J., 1990. Credit and default amongst young adults: An agenda of issues. Journal of Consumer Policy 13, 133-154. Ford, J., 1992. Young adults' use of credit. Paper read at the Seminar on Responsible Lending, Windsor, March. Fox, M., 1992. Student debt and enrollment in graduate and professional school. Applied Economics 24, 669-677. Friedman, M., 1957. A Theory of the Consumption Function. Princeton NJ: Princeton University Press. Holmes, T.H. and R.H. Rahe, 1967. The social readjustment rating scale. Journal of Psychosomatic Research 11, 213-218. Katona, G., 1975. Psychological Economics. New York: Elsevier. Kline, P., 1993. The Handbook of Psychological Testing. London: Routledge. Lea, S.E.G., P. Webley and R.M. Levine, 1993. The economic psychology of consumer debt. Journal of Economic Psychology 14, 85-119. Leigh-Pemberton, R., 1989. Personal credit problems. Bank of England Quarterly 29, 243-245. Livingstone, S. and P.K. Lunt, 1992. Predicting personal debt and debt repayment: Psychological, social and economic determinants. Journal of Economic Psychology 13, 111-134. Maital, S., 1982. Minds, Markets and Money. New York: Basic Books. Rotter, J.B., 1966. Generalised expectancies for internal versus external control of reinforcement. Psychological Monographs 80(1) (Whole no. 609). Sarason, I., J. Johnson and J. Siegel, 1978. Assessing the impact of life changes: Development of the life experiences survey. Journal of Consulting and Clinical Psychology 46, 932-946. Schultz, T.W., 1960. The formation of human capital by education. Journal of Political Economy 68, 571-583. Sullivan, T.A., E. Warren and J.L. Westbrook, 1989. As We Forgive Our Debtors. New York: Oxford University Press. Tokunaga, H., 1993. The use and abuse of consumer credit: Application of psychological theory and research. Journal of Economic Psychology 14, 285-316. Walker, C.M., S.E.G. Lea and P. Webley, 1992. An interview study of the origins of problem debt. Paper read at the conference of the International Association for Research in Economic Psychology, Frankfurt, August.

Student Attitudes to Student Debt

Recommend Documents