Curriculum Choice at A level: why is Business Studies more popular than Economics?
Ray Bachan
Paper presented at the British Educational Research Association Annual Conference, University of Manchester, 16-18 September 2004
Abstract
This paper uses ALIS data to model the factors that influence the choice between Economics and Business StudiesA-level. These subjects are often perceived as close curriculum options and possible substitutes. Subject choice is modelled using an underlying latent variable approach, which employs both a binary and an ordered probit. On the basis of a series of counterfactual exercises an overall average grade differential, a measure of their comparative difficultyin terms of students expected examination performance, is estimated to be 0.7 of an (old) UCAS point. The estimating equation suggests that a unit increase in the grade differential increases the probability of selecting Business Studies over Economics by approximately 16 percentage points. There is evidence that females are less likely to choose Economics over Business Studies, and the more able students, in terms of their average GCSE score and mathematical ability, are more likely to select Economics. There is little evidence of parental background characteristics exerting significant effects on the choice between these two subjects, but there is evidence of ethnic characteristics being significant.
Ray Bachan
University of Brighton
Lewes Road
Brighton
BN2 4AT
United Kingdom
e-mail:
September 2004
Introduction
Concern regarding falling enrolments on certain A-level courses has been frequently expressed in the press and by educationalists in the UK since the late 1980s. With the introduction of the National Curriculum in 1989, falling student numbers may have been expected on some A-level courses. This change was anticipated to impact on those subjects, which were not included in the 14 –16 core curriculum offered at GCSE in many secondary schools and colleges in England and Wales. During the 1990s there was a marked decline in student numbers in some non-core subjects but the inclusion in the core curriculum has not protected other subjects, in particular the ‘hard’mathematical-science orientated A-level subjects (e.g., Physics, Mathematics and Chemistry) from declining A-level enrolments (see, Dearing (1996) and Fitz-Gibbon (1999)). These trends persisted even though there was a rise in participation rates in post-compulsory education during the 1990s
Despite this concern, there has been relatively little systematic examination of the cause of these observed trends, particularly for A-level subjects outside curriculum areas such as languages and maths-science. This is even more surprising for several important reasons. First, it is widely acknowledged that there exists a substantial ‘skills gap’ in the UK labour market and this may in part be explained by curriculum choices at A-level. Second, the A-level examination system in England and Wales has recently been reformed with the introduction of the Advanced Subsidiary (AS) and the Advanced (AS+A2) qualifications in September 2000. The new examination allows students to drop A-level options after the first year of study for which they may qualify for the AS award[1]. Third, recent rate of return studies suggests that there are potential long-term national and personal consequences associated with A-level performance and choice of A-level subject. For instance, Blundell, Dearden, Meghir and Sianesi (1999) report that for males and females the return to A-level qualifications is 11% and 13% respectively. Dolton and Vignoles (2000) detected a wage premium of between 7% - 10% in the labour market for workers with A-level mathematics compared to workers without this qualification[2].
It is against this background that concern over the declining student numbers on A-level Economics has been expressed (see for example Ashworth and Evans (2000, 2001), Bachan and Reilly (2003), and Reilly and Bachan (2004)). The focus of this paper is on the exclusive choice between A-level Economics and A-level Business, which are often cited as closely related A-level curriculum options that offer possible substitutability. Few studies have attempted to explicitly model the choice of Economics over other curriculum options at A-level. The econometric work by Ashworth and Evans (2001) is a notable exception. They find evidence that mathematical ability, prior study of Economics, under achievement in Economics and certain features of the classroom environment are all important factors influencing the decision to select Economics at A-level.
In general research into curriculum choice at A-level finds evidence of several factors influencing subject choice. These include: the previous study of the subject; the student’s perception of the subject’s challenging nature; the likelihood of passing or failing; the interest/enjoyment value of the subject (particularly evident among female students); the type of school (i.e. single sex or mixed); the complementarity between A-level subjects (particularly in the sciences); career aspirations; parental socio-economic characteristics; and the guidance a student receives at school or from parents. Ryrie (1981), Garratt (1985, 1986), McEwan, Curry, and Watson. (1986), Stables and Stables (1995), Gallagher, McEwan, and Knipe (1997), Stables and Wikeley (1997), and Werfhorst, Sullivan, and Cheung (2002), provide evidence on these issues. Knowledge of factors that potentially impact on a student’s choice of A-level is clearly useful to the policy maker and the design of the advanced post-16 curriculum. A primary motivation for this research is to provide new evidence on factors that can influence curriculum choice at A-level, by examining the factors that influence the choice between Economics and Business Studies. It should be noted that a large part of the existing research into the issue of subject choice have been conducted using relatively small samples, and as noted often focuses on A-level subjects within the maths-science and language curriculum.
The Case of Economics and Business Studies
Between 1992 and 2000, the number of students sitting A-level examinations in Economics in England, witnessed a 51% decline, whereas the numbers sitting A-level examinations in Business Studies increased by 80% over the period, see Figure A1 in the appendix. It should be noted that by 1995, the number of students sitting examinations in Business Studies rose above that of Economics for the first time. The pattern described by the data may suggest that, to some extent, students may be substituting A-level Business Studies for A-level Economics (see for example, Hurd, Coates and Anderton (1997))[3]. It should also be noted that over the whole period the total number of students taking both subjects fell by 7%. This may reflect a strong growth in interest in ‘new’ subjects offered by the wider post-16 advanced curriculum offered in the 1990s[4]. Whether some of these students would have selected Economics over Business Studies if these subjects were not available is difficult to discern, but it is a possibility.
Several studies have classified Economics as a ‘hard’ or ‘difficult’ A-level subject (difficult is generally taken to mean ‘severely graded’ i.e. if the grade awarded was generally lower than might have been reasonably expected). These studies provide the robust finding that A-level Economics is relatively more difficult than Business Studies (see Ashworth and Evans (2000) and Reilly and Bachan (2004))[5]. It is argued that the perceived difficulty between these two subjects is an important factor in explaining the formers declining enrolments and the latter’s rise in popularity over the last decade. These observed trends are often taken as a reflection of a rational reaction by students, to move towards the ‘easier’ Business Studies, and away from the ‘harder’ Economics. This choice is assumed to be exercised on the basis of the signals sent to prospective students.
Similar concerns have been expressed in the US[6]. Several studies cite the importance of the comparative difficulty of Economics, and opportunities for substitutability between Economics and other majors (including Business) as factors influencing observed enrolment patterns (see for example, Sabot and Wakeman-Linn (1991), Margo and Siegfried (1996), Bartlett (1995), Salemi (1996) and Salemi and Eubanks (1996))[7].
The methodology adopted allows for the control of a variety of individual, family and school characteristics that are assumed, a priori, to influence a student’s choice between A-level Business Studies and Economics, in one particular year. A measure that standardises comparative subject ‘difficulty’ between the comparator subjectsis constructed, and enters the analysis as an additional regressor in the choice equation. The inclusion of this variable allows for the possible impact that comparative ‘difficulty’ has on the probability of choosing one subject over another. Few studies have used controls for comparative subject difficulty (see Ashworth and Evans (2001) for a recent example), even though its importance has often been noted. The A-level options under study provide examples of a ‘difficult’ and a ‘less challenging’ course of study, that have witnessed opposite fortunes in enrolments in recent times. The results from the analysis presented provide further empirical evidence on factors influencing A-level subject choice. Moreover, it can potentially provide further insights into factors that may have influenced the declining trends in enrolments observed on the ‘hard’ maths-science orientated A-levels.
The structure of this paper is now outlined. The next section describes the dataset used in the analysis, followed by a section containing a description of the econometric methodology employed. The penultimate section reports the empirical results and a final section provides a summary of the conclusions.
Data
The data used are obtained from the A-Level Information System (ALIS) Project administered at the Curriculum, Evaluation and Management Centre (CEM Centre) at DurhamUniversity. The specific data employed in this study are based on performance in the 1998 examinations by a sample of Economics and Business Studies candidates and the information they provided on their personal and family characteristics during their first term of a two year course of study. The sample consists of students aged 16 – 19 years who completed two or more A-levels (excluding General Studies). Once allowance is made for missing values, 2,052 and 3,453 usable observations are available for Economics and Business Studies respectively.
The set of independent variables used and summary statistics are reported in Table A1 of the appendix. They include a measure of A-level performance, prior attainment (average GCSE score[8]), gender, ethnicity, school-type, parental characteristics, examination board, other A-levels studied, the student’s desired occupation and attitudes to the subject. It is important to note that for reasons of confidentiality, the data are limited in a number of important respects. It is not possible to identify either schools or colleges by their names or postcodes and therefore not possible to assign certain factors (e.g. location, funding, staff/pupil ratios, numbers on roll, teacher or class characteristics) to the individual level data used here. In addition, it did not prove possible to identify prior attainment in either GCSE Economics or Business Studies (if taken) for the sample of students. The data set employed in this study has been described elsewhere, see Reilly and Bachan (2004). However, as a preliminary exercise it may be instructive to briefly examine some of the key characteristics of the data that relate to its use in the current context.
The sample of Economics candidates appear better qualified than their Business Studies counterparts using average GCSE performance measures. The average differential in GCSE scores is statistically significant and suggests an average advantage for the Economics sample that is close to one-half of one point[9]. A significantly higher proportion of Economics candidates achieved A/A* grades in Mathematics GCSE relative to Business Studies with smaller proportions obtaining a grade C (or less). The overall chi-squared value confirms that the variation in performance is largely driven by differences in the grade A/A* and grade C categories.
The gender balance for Business Studies is relatively even. The Economics sample has a slight male dominance, but the proportion of female candidates studying Economics in this sample appears on the high side compared to national estimates. For instance, according to the DfEE, 36% of candidates taking A-level Economics in 1998 were female with a comparable estimate for Business Studies of 46% (see DfEE (various issues)).
The parents of students studying A-level Economics are more represented in the professional job classifications than the parents of those studying Business Studies, whilst the latter are more represented in the skilled and unskilled classification. A description of the jobs included within these categories is detailed in the notes to Table A1. These differences are also confirmed by the overall chi-square value for each of these classifications. In terms of parental educational background those students studying Economics have parents educated to a higher level. However, there appears to be few significant differences in terms of parental employment status.
A greater proportion of students following an A-level course in Economics tend to find it ‘more difficult to get down to work’ than those studying Business Studies. But they do appear to think about the subject more than their Business Studies counterparts, perhaps suggesting greater interest in the subject content. It is also interesting to note that a higher proportion of Economics students would like to be employed in the highest professional category, and a greater proportion of Business Studies students would like to be employed in the lower professional and skilled job categories. This difference in proportions is also confirmed by the overall chi-squared value.
The ethnic mix of those studying A-level Economics is more varied than for the sample of Business Studies candidates. It is worth noting that a high proportion of Economics candidates complement their study of Economics with the study of Mathematics and/or Physics at A-level. In terms of A-level performance itself, the aggregate proportion of Business Studies candidates in the B/C categories is about eight points higher compared to Economics candidates. Almost twice the proportion of Economics candidates secure an A-grade in comparison to their Business Studies counterparts but a higher proportion also fail.
Methodology
Subject choice is modelled in two stages using non-overlapping datasets. The sample set of students have either taken a course in A-level Economics or Business Studies but not both. In the first stage, performance equations are estimated for each sub-set. The estimated coefficients are used to construct a predicted grade differential for each individual in the sample by performing a series of counterfactual simulations using the full sample of students. The predicted differential, an estimate of the comparative difficulty between the two subjects under study, enters the second stage of the estimation which models the probability of choosing one subject over the other.
Given the ordinal nature of the final grade classifications at A-level the first stage of the analysis employs an ordered probit. This approach uses an underlying latent dependent variable to model educational performance for each sub-set of students (see Reilly and Bachan (2004) for a discussion on the merits of this approach).
Let yi denote an observable ordinal variable coded 0,1,2,3,4,5 on the basis of A-level performance. Let denote an unobservable variable that captures the performance level of the ith individual. The performance level can be expressed as a function of a vector of explanatory variables (Xi) using the following linear relationship:
+ ui where ui ~ N(0, 2)[1]
Under standard assumptions[10] this model can be expressed in general terms:
Prob[yi = j] = (j – ) – (j-1 – ) for j = 0,…..J.[2]
Where denotes the cumulative distribution function of the standard normal.
The general expression for the log-likelihood function of this particular model is then given by:
L = [wij]loge[(j – ) – (j-1 – )] [3]
where wij = 1 if the ith candidate is in the jth grade category and zero otherwise, and loge is the natural logarithmic operator. Conventional algorithms can be employed to provide maximum likelihood estimates for the parameter vector and the four threshold parameters [1,2,3,4][11].
One econometric problem with the approach adopted is that potential selectivity bias may be introduced into the estimated coefficients. This is of particular concern in the present study as the estimated coefficients are used to construct an average predicted grade differential for each individual in the sample. This problem lies in the extent to which unobservable variables are correlated across the selection and performance equations and is addressed using the procedure outlined by Heckman (1979)[12].
We next simulate how the full sample of students would have performed if they all selected Economics using the estimated coefficients (corrected for selectivity bias) and threshold parameters from the Economics performance equation. This counterfactual can be expressed as:
Prob[Economicsi = j] = ( j = 0,…..J)[4]
where j is jth grade category). A further simulation is performed using the coefficients and thresholds from the Business Studies performance equation, and can be expressed as:
Prob[Business Studiesi = j] = (j = 0,...... J)[5]
On the basis of the predicted grade classification an average UCAS point score is computed for each individual in the sample using expression [4] and [5]. This weighted average is constructed using the old UCAS points tariff as weights[13]. The differential in performance, on the assumption that each individual in the sample took both Business Studies and Economics, can be constructed using the results from the simulations. The differential (Di) can be expressed as:
Di =[6]
where SiB and SiE are the predicted UCAS point score in Business Studies and Economics respectively, and Di can be interpreted as a measure of comparative difficulty between the two comparator subjects. A positive differential implies that a student would have performed better, in terms of the predicted UCAS points, if Business Studies was selected. A negative grade differential implies that a student would have performed better if Economics was chosen. This differential is employed as an additional regressor in the final stage of the analysis.