Fees… an impolite debate.

An interesting, if dispiriting discussion on Irish Economy between Kevin Denny of the University of Kentucky (and associated, as it says on his paper with the Institute for Fiscal Studies) and Joanna Tuffy TD. Denny has written a piece on the impact of the free fees policy at university level.

I’d meant to address this paper because while interesting it struck me as opening up a number of questions. First up Denny restricts his focus to the Universities and from the data deduces that those in low-income groups have not increased their numbers there.

That’s true to a degree, although we’d probably need more data to be clear as to the position today (given that the data used by Denny appears to come from the mid 2000s and therefore may well have changed subsequently).

He argues that:

A priori, given the high level of excess demand (with the number of applicants around twice the number of places) a reform which would stimulate demand seems unlikely to be effective.
Moreover it achieves this by changing the relative price against the population who are under-represented. The reform also clearly generates a deadweight loss: individuals who were willing to pay for an investment now benefitted from the investment at the taxpayers’ expense. An argument in favour of the policy that has been made was that the fees acted as a barrier to individuals whose parents’ income was just above the threshold to qualify for the
Higher Education grant, largely from low income white-collar backgrounds. Raising the threshold would obviously have been a more efficient solution to this problem to the extent that it existed.

One could query the issue of ‘deadweight loss’, particularly given the reality that covenants (which he mentioned) functioned as a means of offering those who nominally were paying fees a means of no entirely circuitously avoiding payment of fees. But I don’t have the costings on that, so it’s difficult to know what the impact was [addendum, here’s some data on that…].

But it’s hard not to feel that the structures once prevailing were remarkable not for their efficiency, but for their complexity, a complexity that concealed considerable transfers from state to individuals in terms of both grants and tax reliefs. It’s far from anecdotal that there were egregious inequalities in terms of the application of both.

And that being the case isn’t there something just a little cosmetic about all this? Because if the we had a system in place that had barriers of entry, however diffuse, above and beyond the issue of exam results (and I’m not ignoring those, but suggesting that their part in this particular aspect of this debate is simply not as important) then surely removal of those barriers – yes at a cost, is in and of itself a positive outcome (obviously from my own perspective I would argue that there should be no barriers at all and that costs should be recouped from general taxation, but curiously that seems to be the least preferred option these days).

Denny points, rightly in part, to the situation in second level as a predicator of success and entry to third level:

On the basis of the results in Table 3 and Figure 1 one expects to find an SES [Socio economic status] gradient with respect to points. Column 1 of Table 4 shows the results of a simple linear regression using the same covariates as in Table 315. What is very striking is the size of the coefficients on father’s SES: the child of a professional can expect to get about 92 points (about .76 of a standard deviation) more Leaving Certificate points than the child of a manual worker. For the children of other non-manual workers the advantage is smaller, but sizeable, at around 47 points. The child of an unemployed man can expect to get about 30 points fewer. This clearly explains why, in Table 3, once one controls for Leaving Certificate points, the direct effect of SES essentially disappears.

But none of this is particularly surprising – surely? The issue of expectation and tradition enters the picture here. If there is no family tradition of attendance at third level then the premium on high levels of attainment at second level will be lower. Particularly if (and I deal with this later) the option of alternative – and more traditional – career paths is available.

The data on the probability of going to university is also of concern, one would think, to almost anyone.

The first three coefficients show the socio-economic gradient: a student whose father is a professional is about 30% (i.e. 30 percentage points) more likely to attend university than one who’s father is semi- or unskilled. For the child of a father in the “other white collar” category the probability of attending university is 11% higher. That such a gradient exists is not surprising since the higher attendance at university by those of higher SES has been documented exhaustively in a series of reports published by the Higher Education Authority and initially compiled by Patrick Clancy starting in 1982, see Clancy (1982,2001) for example and O’Connell et al (2004).

The conclusion that Denny draws is intriguing:

The main purpose of this paper was to investigate the effect, if any, of the abolition of university fees in Ireland in the mid-1990s. This reform, and the possibility of reversing it, has come under renewed discussion in recent years partly because of financial pressures on the universities as well as a desire for greater financial autonomy by them. However a defence of the reform that is commonly offered is that, somehow, it lead to greater access by groups that have been traditionally under-represented in higher education in general and university in particular. On the face of it, this is highly unlikely to be true since prior to the reform many
low income students did not pay fees because they received a means tested grant covering both tuition costs and a contribution to their living expenses. In effect, the reform withdrew the one advantage that low income students had relative to high income students.

Of course the numbers are everything. Denny notes that in UCD prior to free fees 30% of students were in receipt of HE grants. Given the overall societal demography that is clearly a remarkably low number to be drawn from what he terms ‘low income’ groups. In other words the…

Defenders of the policy might to a general increase in access by low income group over-time. This is not serious evidence and ignores, in particular, the secular increase in the supply of places as universities and other higher education institutions expanded, see the numbers cited on page 5 above and Figure 1. Indeed while the absolute numbers from low SES groups rose, as a
share they remained constant17.

But that, to me at least, seems to indicate at least some level of success. Increased numbers from low SES groups, even if as a share of the overall total they remained constant (and the reference he uses dates back to the mid-2000s which would appear to be rather dated), means of necessity that more students from low income groups are experiencing third level at university level and that will provide a process of aculturation that should directly influence their own off-spring and more broadly increase the idea that that third level is a serious option for those from such groups. There’s another point which I am uncertain as to whether he has allowed for, which is that low SES groups are demographically speaking not immutable. They too are subject to growth or reduction in numbers and one could reasonably query what the impact of the economic boom of the past decade and a half had on those numbers and the consequence of that in terms of the numbers entering college from whatever category. One might hazard that in a boom rising wages and greater job opportunities might see some social mobility upwards thereby seeing a shift in numbers from one SES background to another. I don’t know though, to be honest, and haven’t had time to assess the data.

However, taking Higher Education as a whole numbers of those from low-income groups have increased significantly and this emphasis on universities may well be less important than he thinks (or indeed many of the contributors to the Irish Economy thread who seem to see this as explicit proof that free fees have not succeeded in their purpose).

Interestingly the ESRI, as early as 2004, noted the following:

The overall admission rate in higher education in Ireland has increased by an impressive 11% since 1998 and now lies at 55% of the relevant age cohort (17-19 year olds). This improvement has been shared by the various socio-economic groups with the rate for the Skilled Manual Group almost doubling to a range of 50-60% compared to 32% in 1998. The Semi and Unskilled Socio-Economic Group has improved from 23% to between 33-40% over the same period (Table 3.8). At 71%, Sligo is the county with the highest rate of admission to higher education and there has been a 13 percentage point increase in the rate of admission for Dublin (Table 5.4). Seven out of every 10 (68.3%) of those who sat the Leaving Certificate entered some form of higher education (Table 4.4a).

And quotes:

HEA Chairman Michael Kelly comments “This rise in participation is truly remarkable. In the space of a generation, participation has gone from 20% in 1980 to nearly three times that figure in 2004. Impressive progress has been made towards meeting targets set by the Government following the publication of the McNamara Report on Access in 2001 most notably the target for the unskilled group of 33% by 2006 has been surpassed – it is now between 33-40%; the number of mature entrants at 9.4% has almost reached the target of 10% set for 2006.”

In terms of raw data:

In relation to Dublin, there has been significant progress with eight postal districts having a rate above the national average compared to six in 1998. Dublin 1 (North Inner City) has gone up from 8.9% to 22.8%; Dublin 2 (South Inner City) up from 19.5% to 29.5%; Dublin 24 (Tallaght, Firhouse) up from 26.1% to 40%; and Dublin 17 (Priorswood, Darndale) up from 8.4% to 16.7%. Dublin 14 (Rathfarnham, Dundrum, Clonskeagh) at 86.5% has the highest rate of admission with Dublin 10 (Ballyfermot) at 11.7% having the lowest rate.

Even given increased places these figures seem quite convincing as indicative of some shift in terms of the perception of third level in general as a positive path and that the experience of HE at both university and IoT level is now much greater in the socio-economic groups where previously it was not.

And of course, selection of university or IoT may well be down to factors which his research cannot or does not take into account. Proximity to home, cost of attending when the university is far from home, nature of courses and so on and so forth.

Indeed there’s a counter argument that as free fees came into play perceptions of IoT’s improved (not least due to the increasing number of degree course they provided) which may have skewed the situation.

A further point. I wonder if Denny (or indeed others writing in this area recently) is correct that the abolition of fees was centrally about the ‘university’ sector rather than HE in general. The ESRI report above, and indeed Niamh Bhreathnach in various articles, both reference increased participation in HE – not the universities (albeit everyone in this debate appears to use at times the terms university, college and undergraduate fees interchangeably).

Denny references the ‘abolition of undergraduate university fees for Irish and EU students in 1996’. But of course that wasn’t the reform. The reform was the abolition of undergraduate fees for Irish and EU students across the HE sector.

And if one looks at an article Bhreathnach wrote in Public Affairs Ireland in October 2008 she was explicit that the FF/Labour Party Programme for a Partnership Government, published in 1993 promised to assist entrance to higher education, not just universities.

On a further tangent, there’s a serious issue here about the manner in which IoTs are regarded. Personally having externed and taught in both a national institution and in IoTs (up to and including MA level) I have to say that my personal – entirely anecdotal – experience has been of a much stronger and rigorous engagement in the latter in terms of contact hours, project assessments and so forth. Of course that’s not necessarily true in every and all other areas, but it makes me a little dubious about the categorisation of various levels of HE.

There’s another obvious possibility too. As was noted on the Irish Economy thread, given that the Republic went through its most sustained economic boom the likelihood was that many who might otherwise have seen the benefit of third level education went straight into the commercial sector and in particular into areas that they would traditionally have gone to. It’s striking to me, both from my own experience of third level and less directly of PLCs that the uptake in the last two years has been significantly greater than at any time over the last decade. I don’t have a breakdown of income groupings, but – again, admittedly anecdotally – the premium on third level and continued education across all groupings appears to be increasing.

And here is an unashamed anecdote. Twelve years ago I participated in a jobs fair in the community school I attended in my teens. This consisted of sitting at a table in a hall filled with others who had graduated from the school for an evening and discussing the nature of our work with pupils. Almost without exception the attitude when they discovered the wages in the industry after five or ten years, and the [general] necessity to do a degree, was one of disbelief that I, or they, would do that when other more immediate options which would be hardly less remunerative (at least as they saw it) existed. Now that’s not just about fees as a barrier, but it is indicative of an attitude that there were shortcuts one could take at that point in time which would be more profitable, and in truth this mindset has always existed, a sense that one could have wages at 17 or 19 or in ones early 20s when others didn’t. That too is a function of class expectations.

Denny notes that there have been potential other effects ’18 There is anecdotal evidence that this increased the demand for fee paying secondary schools.’.

Well, perhaps, although one presumes that it is not beyond the wit of our political classes to deal with that through taxation or other mechanisms.

Let’s also note that barriers to entry, through means testing and such like, are proven to inhibit take up of place, even when free. There’s a well developed body of research literature that fairly comprehensively demonstrates how targeted/means tested services/benefits are less likely to be taken up by those that need them than universally available ones. Why fees would be any different escapes me.

One aspect of that that always puzzles me is the almost complete lack of understanding amongst some from a middle class background as to how such barriers operate (and it’s interesting how, for example Colm Harmon of UCD, of whom I have good things to say below in terms of his approach seems to be oddly sanguine about how ‘fees don’t matter’ in this piece here). In part I suspect that this is because they tend to have much fewer inhibitions themselves as regards utilising benefits where necessary – for example, the text of the Denny article notes that the covenant system whereby income tax could be set against college fees was used widely, and by definition used by those from better off income groups. That too was a subvention, and one that was closed down when fees were abolished.

Indeed it’s absolutely dispiriting in this debate to have no sense from almost all participating in it on the Irish Economy thread that in working class communities, like the one where I went to primary and community school, there has only been a slow, what I can only term, acculturation to the concept of HE (let alone universities – and as an academic myself teaching in a national institution I’m very very leery about the concentration on the ‘universities’) being a realistic option. The idea that such deeply embedded attitudes can be changed overnight or even within a decade – particularly a decade where the traditional first option of paid employment from the time one left school was never more easily accessible – seems wildly over-optimistic.

Perhaps this is special pleading, but I think it might be interesting to give this just a little longer before we come to any clear conclusions. From my own experience in third level and that of those teaching on PLCs it is evident that since the economic downturn both areas of education are now heavily subscribed as jobs thin out and the alternative of further or improved education gains a premium.

Denny argues though that,

The policy implication that one can draw from this, and a theme that is emerging generally in the literature, is the importance of early interventions: disadvantage sets in early in life. Quite how early is unclear with some authors, notably James Heckman and co-authors, pointing to the necessity of very early interventions in life, see Doyle et al (2009) for an overview. So while this paper demonstrates that performance at the final secondary school exam is the
proximate explanation for the SES gradient with respect to university it is not able to pinpoint the ultimate explanation i.e. what accounts for the differential performance at secondary level. It may be due to the differential effects of secondary schools however this cannot be assumed. It could, for that matter, be due to differences in primary schooling. Students from higher SES backgrounds make greater use of private tuition in preparing for the Leaving Certificate but, paradoxically, it appears to have no beneficial effect, see Smyth (2009).
Perhaps it has nothing to do with schools at all but evolve from the home environment and parental investment in particular. Most likely, it is a combination of these. So while this paper can help rule out some explanations, it cannot pinpoint exactly where the disadvantage sets in. One would need much better data for this.

Indeed, and of course, there must be increased emphasis on primary and secondary levels. But that doesn’t in and of itself mean that the abolition of undergraduate fees was a mistake. What it does imply is that no single policy will deal with all the problematic issues that arise.

It’s interesting to read David McWilliams piece in the Independent where he argues from this research that:

In fact, the abolition of university fees 15 years ago did not improve the chances of poorer children getting into university. The paper also explains why this is the case. There was and still is a shortage of places in university, not a shortage of students. The poorer kids who did manage to get the points were normally on the grant anyway so they didn’t pay in the first place. The middle-class children and their families just got a subsidy for an expense that they were planning to make all along.
The most important thing is how the children do in the Leaving Cert. Poor children do badly in the Leaving full stop. For example, after 15 years of free university it is the case that if your father is a professional, you will, on average, get about 90 points more than if your father is a manual worker.

But it’s only by limiting our focus to the universities (where numbers did in fact go up albeit from all social groups more or less equally) and ignoring the non-university HE sector where numbers from lower-income groups went up significantly that he can make an argument that once more reifies universities over all other parts of the HE. More accurately we could say that the abolition of undergraduate fees clearly did nothing to block the entry of low income SES students to third level. Would I like there to be more people from lower-income groups in universities. I would indeed but I think the overall picture is still better than is being presented here.

But… it also strikes me as an odd argument to suggest that by underscoring or emphasising social distinctiveness through grants and suchlike that one will somehow improve a situation that – given the historical record – had an evident lack of success, across many decades of the fees and grant system, in enabling social mobility and access to third level.

Therefore it seems curious that the current system would be so easily dismissed after little more than a decade in place – and during, arguably, one of the most volatile and anomalous periods in Irish socio-economic history – and that a rising and rather vociferous chorus should be seeking a return, albeit with [usually] unstated modification (although as seen above Denny himself argues that increasing thresholds on HE grants would ameliorate that, though what ‘efficiency’ that measure would bring to the system is questionable), to the status quo ante.

Still, we’ll get no light from this if we are to take the following exchange from the Irish Economy thread as characteristic.

Joanna Tuffy Says: 
May 25th, 2010 at 10:52 pm
The question which should be considered and is not in the paper by Kevin Denny is this: Which socio economic groups were being dettered from going to college by tuition fees?
According to studies by the Higher Education Authority (HEA) between 1980 and 1992, the children of all but three social groups out of 11 improved their participation rate in college. The three social groups were the Lower Professional, Salaried Employees, Intermediate Non-Manual workers – all low-to-middle-income PAYE workers. The participation rate of children from the three social groups concerned not only did not improve but worsened from the years 1986 to 1992. According to the study of entrants in 1998 by the HEA ‘Who went to College in 1998′, that trend was reversed and their participation increased further in the years that followed. Some of those in those three groups would have qualified for a grant and many of them would have been above the income limits for grants.
According to the HEA in its most recent study of all entrants to third level ‘Who Went to College in 2004′ groups that increased participation rates between 1998 and 2004 included “the Skilled Manual Group almost doubling to a range of 50-60% compared to 32% in 1998. The Semi and Unskilled Socio-Economic Group has improved from 23% to between 33-40% over the same period (Table 3.8).”
(Who Went to College in 2004? A National Survey of New Entrants to Higher Education
By Philip O’Connell (ESRI), David Clancy (Fitzpatrick Associates) and Selina McCoy (ESRI). Published by The Higher Education Authority.)
It may well be that the HEA are due to publish a study in 2012 in relation to 2010 college entrants as its pattern has been to carry out these studies every six years and such a study will give a far more comprehensive picture of who goes to college and from what socio economic background,s than Kevin’s paper possibly can. But contrary to what he, and the people who have commented above, seem to have deduced from his paper, the abolition of fees has meant that many socio economic groups that were previously less likely to send their kids to college prior to the abolition of fees, were more likely to do so subsequent to the abolition of fees.
Joanna Tuffy T.D.
Labour

Denny unfortunately does not respond in the most positive way. Indeed – and I say this as someone who works in academia and in research – he displays a curious (that word again) attitude given that this is a matter of public policy and that he’s engaging with a public representative.

@Joanna,I don’t have much to add to Colm Harmon’s. Having said that.. my guess is that you are not familiar with the methodology behind my study and indeed all serious work in this area. Thats why you have completely misunderstood it and its relationship to the HEA reports. Did it not occur to you that you need to know what you are talking about if you are going to contradict, in public, professionals who work in this area & on an important national topic? Would you feel equally free, say, about contradicting an oncologist who produced some research on cancer? Or are we to accept lower standards of evidence? I can only infer that political expediency rather than a desire to understand the truth is behind your comment. Maybe you should consult a taxi driver.
Those HEA reports, which are invaluable & which I cite, do not allow one to identify the effect of the reform. Consequently they do not support your case. Nor do they contradict it. So you are simply unambiguously wrong as a matter of fact in inferring that the HEA reports supports your position. The O’Connell et al piece actually specifically talks about the merits of the kind of multivariate analysis that I carried out. They understand, you do not.

I hold no brief for Tuffy, or the Labour Party – in fact we’ve been fairly critical of them over the years on the CLR. But the very least he could do Tuffy, and indeed all of us others who are not gifted with his insight, the courtesy of explaining in clear terms the methodology that he uses and why he believes it to be correct. He might also explain why he feels the concentration on one sector of HE provides us with an argument sufficient to jettison the concept of free access to HE. That’s pretty basic, but it’s also the right and courteous thing to do.

It is uninspiring to see such a dismissive approach. Tuffy may be wrong, he may be right, or it may be that as with most things that the truth lies somewhere between the two, but he does his case no justice at all by simply wrapping himself in an unexplained cloak of supposed expertise. These are indeed technical matters, but they’re matters that are relatively easily explained.

Moreover, the idea that research in a social science, which indeed my own academic interests lie within albeit in a different area of it, is directly comparable with medical research, particularly when such a comparison is made seemingly in order to respond to dissent is highly unconvincing.

But given that he brings in political expediency, worth noting once more the fact that he is, as it notes on the paper, associated with the Institute of Fiscal Studies. Now no harm there, but it’s worth, as they say, noting that the IFS is regarded by some – no doubt entirely mischievously – as an effectively pro-Conservative think tank – hard to credit for an organisation founded as a direct result of, and in opposition to, the introduction of Capital Gains Tax amongst others by the UK Labour government of the late 1960s – eh? 🙂 ).

And let me give an example of what I consider a more positive discourse… Denny himself references Colm Harmon’s response to Tuffy. Here is that response.

@Joanna
Thoughtful comment but….
1. Denny work is about the University sector – the HEA report covers all HE institutions. The Denny work excluding the IoTs is correct for the task he is undertaking. IoTs students largely faced no fees due to structural fund money (indeed many received a stipend while in HE). So the fee regime change created a discontinuity for the Uni sector which is what Kevin exploits.
2. In absolute terms there are more students, and therefore more low SES students in HE, but we still have a situation where the relative share in University is largely unchanged. That is what the fee regime change was meant to change – it didn’t (and that is true of many countries as is the reverse – introducing does not lower participation – see Oz)
3. Dropouts are much higher in IoTs so who goes to college is a different question to who finishes college!
4. The Denny analysis is not a HEA report. It is a regression model, multivariate, controlling for many factors at the same time. That he finds such strong results is all the more depressing.

That’s the way to do it. That opens up the debate for further questions, some of which I try to engage with above. And that is the way to have a serious discussion free from personalisation or dubious claims of special expertise (by the way kudos to Liam Delaney of Irish Economy for very nuanced moderation and exquisite tact).

And finally, and this is crucial. There’s a notion abroad that there are absolute certainties in all these socio-economic discussions, that logic points to a single sustainable conclusion in all matters. I’m deeply distrustful of that. Indeed I’d suggest that while the data may provide certain results this does not mean that it is unamenable to interpretation. And to think otherwise is – on occasion – as self-deceiving as it self-serving.