Education Notes

Has educational expansion helped to reduce class inequalities in educational attainment?

My overall strategy here is to suggest that educational expansion may not reduce class inequalities in educational attainment. My first set of arguments have to do with the phenomenon, described as a ‘nearly universal pattern’ of waning odds differentials between classes at later transitions. This pattern is often considered strong grounds to believe the expansion of education will achieve the specified effect. I will suggest, firstly, that the effect may be explained by biases in selection. I will also suggest that earlier transition decisions are informed by the potential for later transition decisions. Building on that, I offer an alternative to the common ‘life course perspective’ account of the waning differential, by suggesting that the impact of myopia on different decisions may be a better explanation. Subsequently I go on to accept that, yes, expanded education has generally coincided with a reduction in educational attainment class inequalities, but that this says nothing about causation, and in fact there are plausible confounding variables, and even a story of causation running in the opposite direction to that specified. Following that I say that, even if much of the above is false, once we consider the potential for stratification within levels, a story of inequality may reassert itself. Finally I suggest that measuring educational attainment inequality, even putting aside both the point about within level stratification and broader ethical arguments about whom must be treated equally for education to be equal¹, is trickier than it perhaps seems.

Firstly I want to say a few things about the apparently waning class inequalities at later stages of education. Even if it is the case that odds of completing university are completely independent of origin (as is the case in some nations²), we should recognise that eligibility university is not a randomly distributed phenomenon. The disadvantaged students who reach the stage of eligiblity for tertiary education are likely to, in some ways that have not been fully controlled for, be unusual, since they have avoided discontinuing education at previous points where they could have left education. Even if one could assess the entrance to university odds of one set of disadvantaged candidates eligible for university by controlling in such a way to make that group equivalents of a group of disadvantaged candidates eligible for entrance to an earlier stage, and show that the former group experiences less or no odds disadvantage, may well be missing some unobserved heterogeneity, since working class candidates who get to the stage of being eligible for university are probably unusual in some unexpected or hard to measure way(s).

Secondly, as Raftery and Hout (1993) points out³ if the alleged waning differential pattern obtains, it may be that educational inequality can be ameliorated by expansion, but this may occur only through a compositional effect and not through impacting the odds ratios between classes at any particular stage.

Thirdly, I want to make two related, if both slightly diversionary, points. They are, firstly, that students’ (and perhaps their parents’) decisions about entering one stage of education are conditioned in part on certain beliefs about subsequent stages and secondly, that there is another plausible alternative account of waning differentials that may be better than that provided by the life course perspectives literature⁴.

Consider the following: anecdotally, in the United Kingdom, students who are eligible to attempt A Levels (ie those who have performed satisfactorily at GCSE) do not attempt A Levels unless they expect to successfully attempt some tertiary education. Thus, we might model things as follows. A student attempts A Levels if:

C_X2+X3. < (E_Y2*S_Y2*(1-S_Y3))+(E_Y3*(S_Y2*S_Y3))-E_Y1

Where C = cost (in all senses), X2 = A Levels, X3= Tertiary education, EY1= the earnings from his current level of education EY2 = the earnings from success at A Levels and no further educational success, EY3= the earnings from success at tertiary education, SY2= the odds of success at A Levels, SY3= the odds of success at tertiary education. (Thus, the odds of dropping out between A Levels and university are captured in ‘(1-SY3)’ since dropping out implies no success at uni.

First, let us introduce some ideas about myopia. Assume, given that middle class students’ parents are more able to pass on knowledge about the qualities needed for success at university, there is a myopia differential, such that, even conditioned on odds of eligibility for university, there is more myopia for working class students about this transition. Given that the error in the expected earnings from attempting A Levels is partly a function of the error in the odds of success at university, it is clear that this myopia extends to this earlier transition, such that working class students experience greater mean myopia in assessing whether to attempt A Levels. This is added to by the myopia about the expected success at A Levels themselves, net of the impact thereof on odds of success at university (obviously, failure at a Level implies S_Y3 = 0.) That in itself should not persuade a working class student to not attempt A Levels, since, obviously, the error could be in either direction. However, given we know that there are declining utility returns to additional units of earnings, it is clear that working class students have lower expected utility from attempting A Levels, even if the mean of the estimated odds of...

Buy the full version of these notes or essay plans and more in our Sociology of Developed Societies Notes.