By Stuart Theobald

Scott Burnett has written a provocative column in The South African arguing that fee-free education is not related to quality of education. I really enjoy his columns which poke apart lazy assumptions. He says his latest one is motivated by the fact that some of his “dearest friends” have been expressing concerns that free education leads to lower quality education. I fear this may include me. He also notes the pride we take in robust public debate over policy, though wants this to be guided by logic and evidence and give people space to be heard.

In that spirit I thought I’d respond to his central claim, which I don’t think can be sustained.

First, some clarificatory issues.

Fees and funding are different. I doubt anyone would deny that there is a relationship between funding and quality. That would be a strange argument indeed – absolutely any other measure of quality can be enhanced with more funding. So, better funded universities are going to be higher quality. The relevant question is whether the fact that fees form part of the bundle of funding should affect quality.

One facile answer is “of course it does”. This is because universities in SA have no alternative source of funding other than fees. In order to keep the promise of 0% fee increases made at the end of last year, UCT, for example, has introduced an “austerity programme” that will lead to staff cuts as the university tries to save R120m over the next three years. This will clearly lower quality. All of the other universities in SA will be doing the same. But the argument is not that fees should be eliminated while keeping all other funding equal. Implicit is that fees should be replaced by other funding, specifically tax funding. It is a less facile question whether it is politically feasible to replace fees with central government funding, or whether it would be a good idea, all things considered, but these are not the questions on the table.

Another fairly facile answer is “of course there isn’t”. But this is not saying much if you assume that funding can be provided by the government without restriction. Quality can, of course, be bought at any price. Sweden and Germany are examples where quality universities operate and are free. But no one is arguing that “all fee-free universities are inevitably lower quality” so these counterexamples are of questionable impact (and indeed I’ll argue below the evidence suggests the opposite). In general, no social argument is about necessary relationships (which are better left to the physical sciences). The better claim to refute would be “in SA, the absence of fees would tend to effect quality negatively”. So, my counter argument will be that universities tend to be higher quality if they charge fees.

Another step is needed to focus the question clearly. The claim would have to be that fee-free universities will tend to be at least as good as fee paying universities for a given level of funding. That is the right question to ask. Of course fee-free universities can be high quality at any cost. That is not an interesting point and I don’t think anyone has argued it.

Note that the question here is specifically the link between quality and fees. So it is not relevant to this argument whether charging fees is a good idea all things considered. I have argued extensively elsewhere that tax-funded university education will be regressive, amounting to a transfer of wealth from the poor to the middle class. But none of this has to do with the link between quality and fees.

It is important to note that the question that has been posed is about completely free universities, and not a deferred fees model. So the argument is not about whether universities should be free at point of use, and then recouped over the student’s working life. It is whether students should face any cost related to their university education over their lifetimes. In what follows I leave it open that fees need not be incurred by students at the time of studying. I have argued here why this is preferable for many reasons.

Now, onto the argument.

Imagine two universities:

University A is funded from a government transfer (g), student fees (f), donations, sponsorships and research income (r), and income generated from other sources (o) such as renting out venues or selling paraphernalia. So the university’s revenue is

Ra = ga + fa + ra + ga

University B is funded in all ways the same but instead of student fees, the government grant is large enough to cover the lost student fees, so

Rb = gb + rb + gb

So in order for the issue to be non-facile, the question is the relationship between gb and (ga + fa) when quality is held constant. If it is the case that the tendency is for gb ≤ (ga + fa), then we should conclude that there is no relationship between fees and quality. If, however, it is reasonable to expect that gb > (ga + fa), then there is a link.

Alternatively, we could ask whether it is in fact the case that Ra would = Rb. In other words, should we expect that the revenue of a fee charging university would equal that of a non-fee charging university. When it comes to the available evidence there is more that speaks to this question, and the answer appears to be “no”.

But first, some theory.

George Akerlof’s paper The Market for Lemons: Quality Uncertainty and the Market Mechanism earned him a Nobel prize (shared with Joseph Stiglitz and Michael Spence) for showing why in the second hand car market available supply is likely to be crowded out by low quality vehicles (this was an example for a general model, so don’t get too hung up on the example). Given a market with a mixture of quality cars (peaches) and low quality cars (lemons) you may think consumers would be willing to pay a price that averages between the two, so Pave. The consumer is unable to tell the difference between a peach and a lemon, but the seller is because she has driven the car and been responsible for its maintenance. The problem is that if Pave < Ppeaches then the seller, who knows whether his car is a peach or a lemon, is not going to make it available for sale. As a result, consumers keep encountering cars that are lemons, so the price they are willing to pay falls. This triggers a downward spiral and the market collapses.

This has come to be known as the problem of “asymmetric information” and is now a very vibrant area of study in economics. This body of reasoning suggests that we should expect any market to consist of lower quality offerings, when the person who is paying cannot assess its quality, all else being equal.

There are many examples where this problem exists. Employees whose companies pay their travel costs will fly business class at twice the price of economy even if it’s not worth twice as much to the employee. In nursing homes the recipient of the service is often different to the person who pays. As a result, cases of abuse are not disciplined by the recipient refusing to pay or taking their custom elsewhere. Children at nursery schools may struggle to report to their parents just what quality of education they are receiving, and quality may be worse as a result. Very complex products, like a computer, are hard to assess if you have no experience of computers. Markets come up with ways of dealing with these problems, such as brands, product reviews, standards setting agencies, and so on. But these ways of dealing with the problem inevitably come at a cost that would be unnecessary if consumers were able to exercise an informed decision based on the quality of the goods and services they receive.

When it comes to situations where the government is paying on behalf of citizens who are recipients of goods and services, asymmetric information problems also arise. The SA government finds it notoriously difficult to monitor the quality of primary and secondary schools, for instance. This would be far less of a problem if schools were funded entirely out of fees because parents’ decisions would have a strong disciplinary effect on the schools. Of course, this would be entirely inappropriate, but it illustrates the point. In the absence of the disciplinary effect of the market, government has to implement strong oversight mechanisms, which, in the case of schools, it has so far failed to do.

If we assume students care mostly about the quality of teaching they are going to receive, the likelihood of obtaining a well-paying job after graduation, as well the all-round experience of attending university, they are likely to be better positioned to decide what to study than the government is. A university has an incentive to attract them because they will earn fees. If a university is failing to meet student expectations, fewer will choose it. In the absence of students using their purchase decisions as a disciplinary mechanism to drive universities to meet their wants, how else should we do it? The answer could only be an expensive monitoring mechanism.

So to return to our model of University A and University B and the question of whether gb will be equal to ga + fa. The asymmetric information analysis suggests that it is unlikely to be the case because gb is either going to have to include an additional amount for quality monitoring and assessment of student wants, or total funding will be lower. So

gb > (ga + fa)

Of course, universities save some money in not charging fees because they no longer require the bureaucracy to do so. These savings could be redirected to government level quality monitoring and assessment. The question is which is likely to outweigh the other. Given that universities have to maintain a substantial finance function anyway and collect income from other sources, the savings will be small.

I have isolated one relevant effect, but this is not the only effect we should be concerned about. Having entirely fee free universities would mean no barrier to access for any students. That means the student body would, on average, be brighter and more capable than an alternative situation where some students are dissuaded from studying because of the liability they will occur in doing so (even in a deferred-fee model). This may have a quality effect in that the student cohort of a fee-free university might be superior to a fee paying university. However, there are models that suggest this won’t be the case. Price signalling can be an important effect – so quality students who can obtain funding might still attend a university reputed for being expensive (just like how we’d rather have a handbag that looks expensive). Also, fee charging universities can fund scholarships to cream off the brightest applicants from the fee-free university sector. This appears to be the case in the US. But, while not strictly relevant to the discussion here, it may also be the case that fee free systems entrench a middle class bias in university education. When universities use high school grades or any other aptitude test as a means of access, it is the middle class who can afford to invest sufficiently in their children to ensure they pass such access barriers. This is suggested by the UK experience where enrolment by students from poor backgrounds increased after the introduction of fees. Also universities that rely more on government funding tend to be less autonomous in their own teaching, hiring and research decisions, which has been problem in Germany, which I will discuss further in a moment. Fee free universities also suffer from students taking longer to complete their degrees, another problem in Germany.

So while it is certainly not necessarily the case that fee-free universities are lower quality than fee-charging universities, there are good reasons we should expect that fee-charging universities will be better quality for a given level of funding. This is borne out by the extensive economics literature on consumer and quality relationships. This is widely acknowledged, including in fee-free scenarios. Germany, for instance, has explicit mechanisms to try and give students a direct say over quality measures.

Of course, this argument is at the level of theory rather than the empirical facts. When it comes to the facts, what is notable is that the funding in fee free universities is significantly lower i.e. Ra > Rb. Empirical analysis is difficult, but a comparison of Germany and the UK provides a good start as they represent the two models, with UK universities charging up to £9000 per year in fees and Germany being fee free (largely, there are still some administration expenses incurred by students). A comprehensive comparison of the fee-charging UK and fee free Germany by think tank Higher Education Policy Institute, found the average spending per student is 20% lower than the UK. This differential has opened up since the UK introduced fees. It is striking that the UK’s introduction of fees has led to an increase in funding per student so it was not a case of g being reduced in favour of f. In Germany the opposite has been true – German states have demanded universities increase enrolment, but the average funding per student has been declining.

It is also notable that the UK produces many more graduates, with 48% of those aged 25-34 holding tertiary qualifications compared to 25% in Germany (German degrees tend to take longer, hence number of enrolments can be higher, but graduates fewer). Expanding enrolment is less feasible when there is no automatic accompanying increase in funding which a fees model obviously allows. The expansion in UK enrolment has also been causally linked to the introduction of fees. It is also notable that this UK expansion has disproportionately favoured students from poor backgrounds so the introduction of fees seems to have expanded access rather than reduced it.

Does the UK have higher quality universities than Germany? Rankings are one, albeit problematic, way of answering the question. Here are the number of universities for each country in all three of the main rankings:


QS 2015/2016 Times Higher Education 2015 Shanghai Jiao Tong 2015
UK Germany UK Germany UK Germany
Top 10 4 0 3 0 2 0
Top 100 18 4 11 6 9 4
Top 200 30 11 29 12 20 13
Top 400 47 28 45 28 33 30
Top 500 49 35 37 39


It is clear that the UK dominates at the elite end of these rankings, with Germany becoming more competitive as one moves down them. The base favours Germany with 121 universities, 215 universities of applied science and 56 art and music colleges compared to 140 universities and 11 other degree-granting institutions in the UK. There are 2.5m students in Germany and 2.3m in the UK. What harms Germany’s rankings is that its elite research institutions, such as the Max Plank Institute, do not have students so are not counted as universities. Germany’s ranking would improve substantially if the research output of these institutes was included with its universities. Germany has been investing substantially in improving its performance in these rankings with a programme for excellence focused on its elite universities, specifically designed to push them up the rankings. This is being achieved mainly by pumping a great deal more funding into the targeted institutions, which also have a cap on the number of students.

The Universitas 21 ranking assesses countries higher education sectors as a whole. This ranks the UK eighth and Germany 14th. The United States is ranked first.

Despite the £9000 per year fee cap in the UK, it is notable that 10 times the number of German students study in the UK compared to the number of UK students studying in Germany – 21 237 vs 2 057. That is despite the massive differential in fees. I don’t know why German students choose to incur this major additional cost, but it doesn’t seem unreasonable to assume it reflects their perception of quality.

The comparison between Germany and the UK strongly suggests that where fees are abolished there is typically a reduction in the amount of funding. There are strong suggestions that that is what we can expect in SA, given the tight fiscal constraints on government and the way it has responded to the fees freeze by demanding that universities “come to the party” by reducing their costs. This is the political reality, so I am not at all surprised that South Africans are concerned that an elimination of fees is a threat to quality. But moreover, even if government funding were forthcoming, I have shown here why there are good theoretical grounds to believe it would still lead to a decline in quality. So while I have argued that fees are a good thing, all things considered, I can add that they are a good thing for the sake of quality too.

The solution we need in South Africa is a comprehensive funding plan that will ensure students can access universities while maintaining the incentives to drive quality at universities. Fees are part of the answer. A comprehensive and effective mechanism to fund students is the bigger part.