The influence of financial factors on participation in higher education

Ackaert, Leen; Prof. Verhaeghe Jean Pierre, Centre for Population and Family studies, Department of education, University Ghent

Paper presented at the European Conference on Educational Research, Edinburgh, 20-23 September 2000

ABSTRACT

In this paper we present the results of a study of the relationship between the “higher education participation” and the “financial cost of this participation”. Controlling for the influence of academic and socio-cultural factors, we examined the influence of direct and indirect study costs on the study choice process and the final study decision of students of 18 years old.

At the end of their last secondary school year, 600 pupils and their parents were studied via a questionnaire during the months of May and June 1998. In the beginning of their academic year in higher education, the students were questioned again, this time via telephone.

Our conclusion is that three financial factors - the direct and indirect study costs of higher education and the income of the parents - still have a significant influence on higher education participation. However, these are not the most important ones. Negative attitudes towards school and studying, the idea that higher education is for the rich, and the subjective attraction to quit school, to find a job and to gain money, have a more important negative impact on higher education participation.

1. INTRODUCTION AND DISCUSSION

To increase the education participation, and in particular the participation in higher education, the Belgian government offers scholarships, social facilities and fiscal stimuli to financially weak students. With these subsidies, the government wants to improve the access to higher education and reduce the chance that many youngsters hypothecate their future by not participating in higher education. Does the government meet this target? In other words, do financial factors still have a significant influence on education participation?

Education participation is a quantitative as well as a qualitative matter. It implicates not only an access to higher education for all pupils, whatever their cultural, social and/or financial background, but it also means that students would choose those disciplines which do correspond to their intrinsic capacities and/or interests.

1.1. Which financial factors can play a significant role?

To examine the influence of financial factors on higher education participation, one needs to make a difference between direct and indirect costs.

Direct costs involve expenses which are directly related to the studies. These costs can be obligatory (e.g. books, study materials, transport, accommodation, ...) or voluntary (or social obligatory)(e.g. studytrips, ...). Direct costs can differ from discipline to discipline and from institution to institution (some disciplines imply more expensive study materials than others, some institutions are located further away from home than others,...).

An indirect cost results from the fact that one can not have an income during the study period. Such cost can have a negative effect on higher education participation or on the choice between short term or long term studies. Pupils who fear to repeat an academic year and who do not have the financial means to do so, may start an easier discipline. Or they may quit even during the first academic year. Such fear may already be present during the secondary school years. Pupils who want to quit school after their compulsory school years may have already chosen easier disciplines during these years, even when they have equal academic capacities as those who are not subjected to the indirect cost. Those who come from a lower SES-family have indeed a stronger tendency to choose easier disciplines during their school years. These disciplines imply, however, lower future perspectives (Stinissen, 1987). The relation between SES and study choices have also been found in higher education study choices (Colla, 1996, Groenez 1999).

For some disciplines there is also a third kind of costs that we defined. These costs are related to the future job. Some disciplines demand high installation costs to practise (e.g. dentists or pharmacists). Such costs can have a negative influence on study choices. Although professional installation costs are not directly related to study costs, they play a financial factor in the final study choice.

1.2. Who chooses, when, and how?

Choosing a discipline is the result of a complex process which starts before the end of the secondary school years. The final choice may not be seen as a one-man decision. This decision is taken together with parents and others (peers, teachers, family, ...). Financial factors may interfere such decision. Youngsters can explicitly or implicitly anticipate the financial situation of their parents and choose a discipline which is more modest and less expensive than their own aspirations. They may even quit school.

In addition to financial considerations there are of course other factors which have an effect on the final participation decision. Different academic, social and cultural factors can indeed have an effect on higher education participation. Think of low school grades, repeatedly repeating a school year, negative attitudes towards school and studying, and opinions like “our kind of people doesn’t follow these kind of studies”, “earning money is more important than studying”.

1.3. Research problems and possible solutions

When we want to examine the influence of financial factors, it is very important to differentiate these factors from socio-cultural ones. Frequently there is a strong correlation between socio-cultural factors and financial factors. Even when we find a clear relation between higher education participation and financial factors, we have to realise that this relation may partly be the result of underlying socio-cultural factors. The latter may even be the main cause of non-participation.

For this reason, when we want to examine the influence of financial contributions to higher education participation, we have to control for these social and cultural factors.

Dissociation of both factors (financial and socio-cultural) can be realised in a direct or indirect way. We can simply ask pupils and their parents if financial considerations have played a role in their decision. However, social desirability may mask the real answer. The indirect way is to include different financial and socio-cultural variables in the same model.

Financial factors are:

- knowledge of scholarships and social facilities,

- family income,

- subjective estimation of the study cost,

- the chances to find a job, and

- the expected earnings when the youngster would leave school.

The subjective estimation of the study cost is based on the assumption that a high estimation leads to a decision not to choose a particular study. Not choosing a high estimated discipline can be an indication of a financial influence. The estimation of the chances to find a job and the expected earnings is a tool to measure the indirect study cost. Both factors are based on the assumption that the higher the job chances and earnings are estimated, the more attractive it is to quit school and the higher the financial losses are estimated when one starts studying.

Socio-cultural factors are a variety of attitudes towards school and studying, and include perceptions of the relation between studying, social background and earning money.

2. RESEARCH QUESTION

Based on the above discussion we formulated the following research questions:

  • Is there an SES difference between youngsters who don’t participate in higher education, youngsters who go to university and youngsters who go to college?
  • Is there an academic degree difference between those three groups of youngsters?
  • Is there a social and cultural difference between quitters, university students and college students?
  • Do these three groups have another estimation of the direct and indirect study costs, and do they have a different knowledge of scholarships and other social facilities?

PUPIL QUESTIONNAIRE / PARENTS QUESTIONNAIRE MAY–JUNE 1998

INDEPENDENT

FOLLOW UP INTERVIEW BY PHONE NOVEMBER 1998

DEPENDENT

3. DATA COLLECTION AND SAMPLE DESCRIPTION

3.1. Data collection

At the end of their last secondary school year, 629 pupils were studied via a questionnaire during the months of May and June 1998. At school, each pupil also received a questionnaire for his/her parents. They were sent back to us by mail.

In the beginning of their academic year in higher education, the same students were questioned again, this time via telephone. We asked them if they quit school or if they started their intended higher education discipline in college or university.

3.2. Sample description

At school level, the sample is representative on different school criteria, like region, grade and private or public schools. On the pupil level we have an over response of the higher grades and on the parent level we have an under response of parents with a low education. Because we are interested in the influence of financial factors on higher education participation we weighted the data based on the national figures of the education distribution.

4. THE DEPENDENT AND INDEPENDENT VARIABLES

4.1. Dependent variable

Originally, one of the purposes of our research was to investigate whether or not financial factors have an influence on the discipline that students choose. Because of the low observed frequencies in the different disciplines, we were obliged to reformulate our research question and recode our dependent variable in three main categories: 1) quitting school; 2) following college courses; or 3) following courses at a university. The weighted frequencies of our dependent variable are presented in table 1.

Dependent variable / Absolute frequencies / Valid percent
quitting school / 44 / 8,4%
college / 346 / 66,4%
university / 131 / 25,1%
valid total
missing
Total / 521
108
629 / 100%

Table 1: Frequency distribution of the dependent variable

Most of the pupils do not quit school and follow courses at college (66,4%) or university level (25,1%). Only 8,4% of the questioned youngsters quit school.

4.2. Independent variables

4.2.1. Academic

Originally:

- the %-chance you give yourself to succeed in higher education (0% - 100%)

- “I have enough capacities to succeed in higher education” (1 - 5)

- amount of failures in secondary education (0 - 3)

- type of secondary education (dummy)

After data reduction: (principal component analysis)

QAR_acad: 39% explained variance (1 = Q1 - 4 = Q4 of the factor scores)

- the %-chance you give yourself to succeed in higher education (factor score = 0,659)

- “I have enough capacities to succeed in higher education” (factor score = 0,753)

- amount of failures in secondary education (factor score = - 0,529)

- type of secondary education (factor score = - 0,547)

4.2.2. Social-economic background of the parents

Originally:

- age of the parents

- family income

- education of the parents

- unemployment frequencies of the parents

- amount of employed parents

After data reduction: (principal component analysis)

QAR_emppar: 47 % explained variance (1 =Q1 - 4 = Q4 of the factor scores

- unemployment frequency of the father (factor score = 0,584)

- unemployment frequency of the mother (factor score = 0,754)

- amount of employed parents (factor score = - 0, 727)

QAR_edupar: 81 % explained variance (1 = Q1 - 4 = Q4 of the factor scores)

- education of the father (factor score = 0,963)

- education of the mother (factor score = 0,876)

QAR_agepar: (1 = Q1 - 4 = Q4 of the original age scores)

QAR_incpar: (1 = Q1 - 4 = Q4 of the original income scores)

4.2.3. Socio-cultural factors

Originally: attitudes of the pupil

- negative attitudes towards studying: studying is a waste of time (factor score = 0,752), studying is not useful for a future professional career (factor score = 0,730), studying is useless (factor score = 0,709), a lot of things are more important than studying (factor score = 0,554), higher education gives a higher chance to a career (factor score = - 0,501), higher education is not interesting (factor score = 0,416)

- positive attitudes towards school: I like to go to school (factor score = 0,853), I feel at ease at school (factor score = 0,785), I’m bored at school (factor score = - 0,772), I like to study (factor score = 0,598)

- higher education is something for a higher social class: studying is something for rich people (factor score = 0,788), higher education is nothing for people of my own social class (factor score = 0,760)

After data reduction: (principal component analysis)

QAR_attpup: 48% explained variance (1 = Q1 - 4 = Q4 of the factor scores)

- negative attitudes towards studying (factor score = 0,787)

- positive attitudes towards school (factor score = - 0,736)

- studying is something for a higher social class (factor score = 0,554)

Originally: attitudes of the parents

- studying is something for a higher social class: higher education is something for rich people (factor score = 0,755), we don’t have enough financial means to finance higher education (factor score = 0,743), higher education is nothing for people of my own social class (factor score = 0,472), studying is useless (factor score = 0,366), studying is not useful for a future professional career (factor score = 0, 366)

- negative attitudes towards studying: higher education is not interesting (factor score = 0,747), studying is a waste of time (factor score = 0,570)

- earning is more important than studying: a lot of things are more important than studying (factor score = 0,825), earning money is more important than studying (factor score = 0,662)

After data reduction: (principal component analysis)

QAR_attpar: 51% (1 = Q1 - 4 = Q4 of the factor scores)

- studying is something for a higher social class (factor score = 0,773)

- negative attitudes towards studying (factor score = 0,758)

- earning money is more important than studying (factor score = 0,617)

4.2.4. Perceptions of direct and indirect costs

Originally: perceptions of the pupil

- expected monthly study costs

- expected job chances after secondary education (0% - 100%)

- expected wage after secondary education

After data reduction: (principal component analysis)

QAR_jobpup: 67% explained variance (1 = Q1 - 4 = Q4 of the factor scores)

- expected job chances after secondary school (factor score = 0,823)

- expected wage after secondary education (factor score = 0,823)

QAR_stcpup: (1 = Q1 - 4 = Q4 of the original cost scores)

Originally: perceptions of the parents

- expected monthly study costs

- expected job chances after secondary education (0% - 100%)

- expected wage after secondary education

After data reduction: (principal component analysis)

QAR_jobpar: 66% explained variance (1 = Q1 - 4 = Q4 of the factor scores)

- expected job chances after secondary school (factor score = 0,819)

- expected wage after secondary education (factor score = 0,819)

QAR_stcpar: (1 = Q1 - 4 = Q4 of the original cost scores)

4.2.5. Knowledge of financial support (subsidies)

Originally: knowledge of the pupil

- amount of correct answers on 16 questions about financial support

After data reduction:

QAR_knpup: (1 = Q1 - 4 = Q4 of the original knowledge scores)

Originally: knowledge of the parents

- amount of correct answers on 16 questions about financial support

After data reduction:

QAR_knpar: (1 = Q1 - 4 = Q4 of the original knowledge scores)

5. ANALYSIS TECHNIQUE

Because our dependent variable is a categorical variable and our independent variables are quantitative variables, we use multinomial logistic regression to analyse our data. The purpose of this technique is to explain the natural logarithm of the probability of one category versus the probability of the other category. By comparison with discriminant analysis, multi-nomial logistic regression analysis demands less severe data properties and the expected values, based on the regression equation, are more in accordance with the observed values. A mayor disadvantage of this technique is the necessity of a considerable survey size. When we compose a model with a lot of independent variables in the regression equation, we increase the probability of empty cells, and as a consequence the results are less reliable. For this reason we fit different models, we reduce the amount of independent variables in the regression equation and we recode the original or the factor scores into categories of a larger size.

The different models are:

The pupil model:

ln(quitting school / college / university) = QAR_stcpup + QAR_jobpup + QAR_acad + QAR_attpup + QAR_knpup

The parent model:

ln(quitting school / college / university) = QAR_stcpar + QAR_jobpar + QAR_attpar + QAR_knpar

The socio economical model:

ln(quitting school / college / university) = QAR_empar + QAR_edupar + QAR_agepar + QAR_incpar

Finally we composed a new model, which is the merged model. It includes the significant effects selected from the above models. The regression coefficients in the output are not standardised. Normally we cannot compare these regression coefficients. But because the scores of the independent variables are all quartiles, comparison is allowed. The analysis output contains two parts. The first output explains the choice between quitting school (1) and college (0). The second output explains the choice between college (1) and university (0).

6. ANALYSIS RESULTS

6.1. The pupil, parent and socio economical model

In table 2, we find the results of the different models. All the models have a significant chi-square. For an easier interpretation of the figures we analysed the models separately for each dependent variable, that is quitting school versus college and college versus university.

Quitting (1) college (0) / Regression coefficient / Wald-statistic / df / Significant level
Pupil model / QAR_stcpup / 0,651 / 5,398 / 1 / 0,020
QAR_jobpup / 0,885 / 8,563 / 1 / 0,003
QAR_acad / -0,694 / 4,683 / 1 / 0,030
QAR_attpup / 1,110 / 9,662 / 1 / 0,002
QAR_knpup / -0,473 / 3,167 / 1 / 0,075
Parent model / QAR_stcpar / 0,368 / 0,777 / 1 / 0,378
QAR_jobpar / 0,088 / 0,071 / 1 / 0,789
QAR_attpar / -0,121 / 0,118 / 1 / 0,731
QAR_knpar / -0,238 / 0,357 / 1 / 0,550
Socio economical model / QAR_emppar / -0,005 / 0,001 / 1 / 0,977
QAR_edupar / -0,003 / 0,000 / 1 / 0,991
QAR_agepar / -0,730 / 3,076 / 1 / 0,079
QAR_incpar / -0,639 / 5,798 / 1 / 0,016
College (1) University (0) / Regression coefficient / Wald-statistic / df / Significant level
Pupil model / QAR_stcpup / -0,142 / 0,826 / 1 / 0,364
QAR_jobpup / 0,241 / 3,011 / 1 / 0,083
QAR_acad / -0,161 / 1,175 / 1 / 0,278
QAR_attpup / 0,037 / 0,062 / 1 / 0,804
QAR_knpup / -0,474 / 9,838 / 1 / 0,002
Parent model / QAR_stcpar / -0,535 / 9,948 / 1 / 0,002
QAR_jobpar / 0,119 / 0,819 / 1 / 0,365
QAR_attpar / 0,376 / 6,880 / 1 / 0,009
QAR_knpar / -0,359 / 4,654 / 1 / 0,031
Socio economical model / QAR_emppar / -0,017 / 0,026 / 1 / 0,873
QAR_edupar / -0,406 / 8,271 / 1 / 0,004
QAR_agepar / -0,092 / 0,107 / 1 / 0,743
QAR_incpar / -0,251 / 2,596 / 1 / 0,107

Table 2: multinomial logistic regression coefficients of the pupil, parent and socio economical model

6.1.1. The pupil model

Our model is significant (model fitting: chi-square = 17,901, df = 5, Sig. = 0,003). Each of the independent variables in the pupil model contributes significantly to the explanation of the dependent variable. However, the independent variables do not explain the same dependent probabilities.

When we compare the “quitting versus college model” with the “college versus university model” we notice that there is quite a difference in the amount of independent variables with a significant effect on the dependent variable. While quitting school is much more a result of factors which are related to the pupil, the choice between university or college is only the result of the pupil’s knowledge of financial support and students’ live. Attitudes which are related to school and studying, financial perception of study costs, expected wage or job opportunities and academic capacities do not contribute to the explanation of the choice between college or university.

Quitting school is the result of a lot of factors which are related to the pupil. Negative attitudes towards studying and school, and the idea that higher education is for the upper-class, lead to quitting school. Also financial and academic factors do play a role in the explanation of quitting school. When a pupil expects high education costs and beliefs that he/she will find a job and earn a relative high wage after secondary education, he/she will more easily quit school. Pupils who followed a technical discipline and failed during their school years will more likely quit school. After verifying all these effects we cannot say that pupils who quit school have less knowledge of financial support than pupils who follow college.