Is higher education looking too much like banking in 2007?

In some important ways, English universities are behaving like banks. They are selling big loans to students, almost half of whom will not be able to fully pay them back. But like the pre-2008 banks, universities do not face the moral hazard of their bad debts. The rewards of having more students are enjoyed by university staff. Their numbers grew last year to more than 380,000, in spite of a 6% drop in students. But when students do not pay back their loans, the cost falls to the taxpayer. And of course most taxpayers have not been to university themselves, but are required to pay for others to have this experience. When many of those taxpayers are poorer than the former students, it is a regressive tax on them. Whether it is students or taxpayers who foot the bill, there is a pressing case for more Government intervention to protect them both against poor decisions in taking on these financial liabilities. Right now, it is clear that too many students are attending universities or following courses which have unsatisfactory completion rates, offer limited chances of graduate level employment and cost students more than they gain financially. And as this means that they won’t be able to pay back their loans, the bill for these particular students falls to the taxpayer. 

But wasn’t the (highly controversial) student loan system meant to shift the full cost of a university education onto the students who receive it and weren’t they then going to decide whether it was worth it or not? Well, “Yes” and “No”. One key issue is that repayment is linked to future earnings. On the “Yes” side, someone who earns enough over a 30 year period after they complete their studies will repay their loan in full. So, for example, doctors who work continuously are likely to repay their loan after 25 years. But on the “No” side, a typical teacher who works for 30 years will only every repay about half of their loan. Indeed, even the doctor will not repay all the costs of their training, as their fees are subsidised by the government. When the new loans system was brought in, the Government estimated that 35% of money lent would not be repaid. But recently, the National Audit Office has suggested this is likely to be at least 50%. On top of this level of non-payment, there are substantial subsidies for the poorest students who receive maintenance grants and for higher cost courses. All in all, it means that the taxpayer will be paying more for university education than it did before the loans were introduced. In these terms, it is hard to believe how much political capital has been spent on the loans controversy. Armed with these facts, I suspect that political history will label the introduction of student loans as one of the presentational failures of the century. Income-contingent repayment is a brilliant innovation. But rather than branding it a debt and stressing that all students will end their studies having to payback £40-£50,000 (which has cast such a negative pall) surely it would be better to say that the taxpayer will cover the full cost of the education, unless students earn enough to start paying it back. This could be made concrete through a simple graduate tax, with a percentage premium on the basic rate  of income tax for those who pay it and then a further percentage premium on the higher rate for the highest earning graduates. The obligation to pay such a tax could be time-limited, as in the current 30 years. Such a graduate tax could be constructed so that the full costs of university education fell on those who benefited from it, not on the, by definition, poorer taxpayers who were not lucky enough to have the experience themselves. A variant would be to share the graduate tax between employees and employers, e.g by putting it on National Insurance rates rather than income tax as these are paid by both sides. There is clearly a logic to asking employers to pay for the benefits they gain. These options have the advantages of fairness and transparency (those who benefitted pay all the costs, those who can pay more do so, but those who didn’t have the experience don’t pay at all). However, it could remove the incentive for universities to innovate and / or offer different prices for courses. One way to tackle this would be to allow variation in fees, but convert this into a differential and personal premium on income tax or National Insurance for the individual. So those who spent more would sign-up to a higher tax rate. 

However, whilst this new approach has benefits, it does not tackle the level of failure in the system. To use the financial analogy, too many sub-prime loans will still be made (courses which predictably lead to high levels of default) but this time the cost will borne by other customers (graduates who will pay a higher level of tax to cover the costs of the bad debts) rather than general taxpayers. Of course, the equivalent of the bank employees and shareholders (university staff) still bear no personal risk – they take their income up front. If we want to reduce the level of failure, it’s worth looking at existing “lending” in some detail. 

Now, if universities were a financial institution this sale of a credit product would now be strictly regulated. The university would be expected to advise the individual on whether the product really suited their needs, its previous performance and the financial returns versus costs. If they didn’t do this effectively, they would be guilty of mis-sellling. As the guarantor of hundreds of billions in loans (the NAO says unpaid loans will total £200 billion in 30 years time), one would expect a government to clamp down on irresponsible lending, i.e. selling courses to students where completion rates are poor, graduate job prospects are low and future earnings may be no better or even worse than not going to university. If one applied this analogy from the financial services industry, then there would be an arguable case that some universities are mis-selling to individual students and governments are failing to protect the taxpayer from massive liabilities. 

A typical student loan for an English student will now total £43,000, including fees and maintenance. The average student expects to have other debts as well (up to £10,000). But they will also have foregone a salary for 3 years as well. If we assume that this was £19,000 per year, then they will have foregone £57,000. After tax on income, this means that a degree costs a student about £100,000. Against this cost the proponents of higher education like to cite the extra earning power of graduates. The Government says this is £168,000 for men and £253,000 for women. (The women’s figure is so much higher as non-graduate women earn such low wages). These figures suggest that a university education more than pays for itself. But these earnings figures are a cruel deception for many – they are an average which masks that some degrees (e.g. medicine) have an earnings premium of more than double the average, whilst many subjects have a very low premium and indeed some result in making graduates worse off than those who didn’t go to university at all. For those who want their degree to be profitable, there is a rhyming couple of acronyms to remember – STEM (Science, Technology, Engineering, Maths) and LEM (Law, Economics, Management). These are the subjects which really boost future earnings. By contrast, the arts and humanities produce a poor return in terms of future income. But the problem for many students is much worse than this. A fifth of students don’t complete their chosen degree; many do not get the 2:1 or 1st which gives access to better jobs; a university education was not appropriate for a significant number of students. 

This is not about academic snobbery. Many of the newer universities offer very good value to their students. And many of the more prestigious universities offer individual subjects which deliver the student and the taxpayer poor value. Some of these subjects are well established and sound like a good bet to prospective students. A classic example is English. This is a popular subject, with 60,000 students across 100 universities. But only half of them will end up with a graduate level job. I can speak from the heart on this subject. I did a degree in English. It was a relaxing way to spend 3 years – savouring fat novels, pondering pious poems and learning enough Anglo Saxon to decipher fragments of monastic propaganda. But it was utterly lacking in intellectual rigour, vocational preparation or engagement with the contemporary world. And that was at the country’s top university! For me the degree was preceded and succeeded by a love of literature. In the same way that it was preceded and succeeded by a love of Premier League football, hillwalking and American TV drama. I was about to say and “But thankfully the taxpayer doesn’t pay for people to study those subjects…” when I realised that in fact the taxpayer does just that. The point is that we don’t need university departments funded in large part by the taxpayer in order to have a vibrant culture in the arts and humanities. Many of the most voracious consumers of the arts studied science and engineering; many of our top historians are not academics; the best political thinking comes from think tanks not universities; the most important reviews of the arts come from specialist journalists, not academics. 

If we are to protect students and taxpayers from wasting time, effort and money, then we draw some lines in the sand about which universities should be attended and which courses should be followed. One way of doing this would be to say that universities will only be accredited for student loans if they achieve a range of minimum standards:

(1) Completion rates – The very best universities have a drop-out rate of less than 1.5%. The worst have drop-out rates approaching half of all students failing to complete the degree they start, with more than 20% dropping out in year 1. It would seem reasonable to say that no state funded university should have a drop-out rate which is more than 10 times worse than the best – i.e. 15%. But if we set the minimum standard at this rate (i.e. 85% of those starting a degree should finish it), that would disqualify more than one-third of UK universities. We could consider a lower rate, e.g. if we tolerated a drop-out rate of 20%, then this would still disqualify 20 out of 121 UK universities (17%). 

(2) Job prospects – It seems reasonable for students  to expect that more than half of the students on their course will get a professional or graduate level job. But this standard is not met in many cases. The very best universities get almost 90% of their students into this level of job. But if we set the minimum standard at 55%, then 20 out of 121 UK universities would fail the test. If we set the standard at 60%, then a third (40) would fail this test. 

(3) Entry qualifications – The standard of education achieved before university is critical to both success at university and in future employment. The top university requires over 600 UCAS points. It seems reasonable to set a minimum entry requirement at half this  level, i.e. 300 points. But more than a quarter of UK universities would fail to meet this standard. 

(4) Quality of degrees awarded – The proportion of degrees classified as 1st class or 2:1 is a good measure of relative quality, as standards are moderated across universities. It is also critical to getting a graduate level job, as many employers specify it as a minimum qualification. The earnings implications are clear – getting less than a 2:1 costs the average student £80,000 in lost future income – almost the same again as the cost of doing the degree. The top university achieves 90% at 1st or 2:1 ; the bottom university achieves 45%. If we set the minimum standard at 60% of students getting a 1st or 2:1, then 29 out of 121 universities (a quarter) would fail the standard. 

So there is a choice here about the level of challenge to be set. If we set the minimum standards at a high level, then one-third, some 40 universities, would not currently meet the standards. If we set an unchallenging test, then 20 universities (about 1 in 6) would currently fail to meet the standard. If we said that a failure to meet any one of these standards disqualified a university, the numbers failing the tests would be higher (as some fail on one, but pass others). These minimum standards could be set by university (so that they could have some subjects below the standard, so long as they were balanced out by others above it) or for each individual course. The logic suggests that every course has to meet the standards – otherwise, large numbers of students would still get something below the minimum. Why would that be fair or acceptable?

In spite of tightening up on the quality of education which is supported by student loans, there would still be a proportion of the student loans which are not repaid because people do not earn enough during their career (e.g. school teachers). This cost falls to the taxpayer – whether that is the graduate-tax payer or the general taxpayer. The level of this subsidy is an important and material fiscal question, which will no doubt be revisited in coming years. However, whoever pays (student or taxpayer) we should be looking at reducing the cost of undertaking a degree. In doing so, it would be a false economy to cut academic pay or to reduce the facilities available to students. But there are two big opportunities to reduce cost. The first is to reduce the time taken to complete a degree. For most undergraduates, they only attend university for half of the year. It is hard to see why a first degree could not be completed in 2 years, rather than 3 years. That would still leave students with quarter of their time on leave. Whilst it may not reduce tuition fees by that much, assuming the need for similar amounts of tuition, it would reduce the typical student costs for a degree by at least a quarter (given that they would have one less year of maintenance and one more year of earning a wage). The other game changer on costs must lie in technology. At some point in the next 5-10 years, there will be a complete digital revolution in our universities. It’s hard to see how teaching, learning or research will not transformed by technology from top to bottom.  The change has clearly started, but the acceleration has not yet reached the escape velocity needed from the gravitational attraction of a traditional education. I am not saying that we all we need is global websites full of MOOCs – the fact that only 7% of them get completed tells us something about the need for a more structured mix of learning. Using its student loan leverage, Government could seek to reduce the cost of degrees by saying that it’s funding assumptions are based on undergraduate degrees lasting 2 years and online content being a certain percentage of the teaching and independent study in degree courses. Even on a targeted basis, applying these criteria to just the courses which have lower economic returns would reduce the debts faced by students and taxpayers, whilst improving the value for money for all involved.

I know these arguments won’t be popular with everyone, especially many in universities. It’s easy to retort that education should not be measured in such utilitarian ways – that we need the arts and humanities; that we need to offer the chance of higher education to everyone who wants it; that this is turning the clock back to a time when fewer people went to university. If universities and current degrees were the only education on offer then I would concede these points. Indeed, one of my favourite quotes from one of my favourite philosophers, John Dewey, is that “Education is not preparation for life; it is life itself”. But as Dewey implies, firstly, education comes from many sources – from our work, our training, our private study, our curiosity, our reading, our travel, our friends. Secondly, not all education is an investment, much of it is consumption. (My ongoing self-taught science being a good example). Thirdly, if the minimum criteria were put in place, universities would fight back with new degrees that met the standards required. But what is clear that is education has a cost – it takes up our time and the people who provide it usually want paying. And if the great financial crisis has taught us anything it must be that someone always has to pay – whether it is those who borrow, those whose lending goes bad or taxpayers who pick the tab for the other two. The same is true of education. But, to use the banking analogy,  there is clearly a need for Government intervention to make sure that lending is responsible, that repayment is fair, that costs are kept to a minimum and that consumers triumph over suppliers. 


Is this the perfect 2015 Election promise?

I have a simple, single sentence promise to be adopted by whichever politician wants to win the General Election in 2015. I suspect that by transforming the lives of most households in the UK within 2 years of the Election, this promise might win the 2020 Election too!

The perfect 2015 Election promise needs to tick the following boxes:

– Address the cost of living crisis, especially that faced by hard working, but low income families

– Capture the imagination of women voters, especially those aged 30-45.

– Show that the political party is on the side of ordinary families against vested interests

– Reverse the view that the next generation will not have it as good as their parents

– Produce a trump card that shows we know how to win the global economic race 

– Be quick to happen, with a simple lever pulled in Whitehall, and most households having their life transformed within 2 years. 

– Not cost too much, but where it does cost more that money needs to be an obvious investment in the future

– Prove that politicians can do big things that matter in the real world, and quickly. 

A daunting checklist – but there is an answer staring us in the face. What about this for a simple manifesto promise – “From September 2016, all state funded schools will, by law, provide 45 hours of education per week for 45 weeks of the year”. This increase by two-thirds in the time that kids spend at school is designed to allow all parents to work full-time without the need for additional childcare. The average employment leave would cover all school holidays. The average working day would give most parents the chance to do a full time job, in between dropping off and picking up their kids. It introduces the length of day and school year which has been shown to have dramatic results on kids’ education in the US. It gives teachers the same sort of working week and annual holidays as other hard working professionals. It’s disruptive enough to be a real game changer in education, in employment and the economy more generally. 

We will come to the education arguments (and the teacher fury) in a moment. But the role schools play in our national and family life is far too important to leave to teachers. And it’s certainly too important to leave to their knee-jerk, as opposed to thoughtful, responses. So let’s focus on some of the non-educational arguments first. Top of the list is allowing more women to work and more women who work to work more. Women working is vital to our economic growth. There are estimated to be at least 1m more women who would work if childcare was easier. That’s a big boost to our workforce, not just in quantity, but critically in quality given the numbers of highly skilled women not working, or not working full time. Four out of ten stay-at-home mums want to work and a fifth of mums who work want to do more hours. Only just over half of single mothers work, compared to nearly three-quarters of mums with partners. 70% of working mothers earn the same as the father or more. This is a big issue for lower middle income (LMI) families (those whose situation is superbly covered by the always excellent Resolution Foundation). Over the last few decades, male breadwinners in LMI families have brought home less and income growth has depended on in-work benefits from the State and female employment. We have passed the high-water mark of what benefits can be expected to provide. So we need to focus on boosting female employment. This grew strongly in the 1970s and 1980s, but went nowhere in the 1990s and 2000’s.  Women are now half of the workforce, but 42% are part-time, compared to just 12% of men. Part-time women earn on average £8.12 per hour (vs full time on£12.00), but this means that 50% earn less than £8.12. As well as tackling low wages (see my post on the Minimum Wage), the best thing we can do to relieve poverty and the cost of living crisis is to help more women work more. Two-thirds of women say that affordable childcare is the biggest barrier, with 40% citing it as the main barrier. Full-time school should provide all the free childcare that most people could want. Imagine if, of the 5m women working part-time for an average of 18 hours per week, 20% worked 10 hours more (as they tell pollsters that they want) and 1m mums who don’t work also started working, with half being full time and half doing 18 hours per week. That’s 2m mums working more, 1m full time equivalent extra workers. This would give the UK the same sort of female employment rates as the Scandanavian countries. The State would save money – it wouldn’t have to pay parents for school age childcare costs (through all its existing vouchers and credits) and of course, it would have the tax revenues of the new workers, it could expect parents on benefits to work more. The non-educational benefits go much further. Just one example is telling. 30% of all youth offending happens between 3 and 6pm each day, in the period between school finishing and parents getting home. Full time school would eliminate this period, the peak period of youth offending. 

But is it really a good idea for kids to spend so much time at school? The key answer to that question lies in what they do with the time at school. But first, let’s check the numbers. The 45/45 school year equals 2025 hours. 45 hours is 9 to 6, or 830 to 530. Assuming 8 hours sleep each day, a kid has 5,840 waking hours each year. That means kids would still only be at school for about a third of their waking hours. Put that way, doesn’t it seem half-hearted that they currently only spend about a fifth of their waking hours at school? A two-thirds increase in education still only equals a third of a child’s time. So in numerical terms, it is clearly not too much. What about the quality of the experience? We know that the current curriculum is overloaded. In most Western countries there has been huge change in the curriculum in the last couple of decades, but no increase in the time  available. The UK positives are that more time is spent on English, Maths, assessment and (for older kids) vocational subjects. But in the zero sum curriculum, time spent on other subjects has declined to make room. The shortage of teaching hours also means that the teaching moves too quickly – there is lots to cover, not much time to explore it deeply and rarely time to help kids play catch up if they didn’t get it first time. Teachers are stressed, children are rushed and learning is frustrated. We also know that the long school holidays impact badly on kids’ progression. This impacts most badly on poorer kids. Malcolm Gladwell showed, in Outliers, that poor kids make the same progress as better-off kids during the school year, but they stagnate during the long holidays, whilst better-off kids continue to progress. The main advantage enjoyed by better-off kids in education is that they need school less than poorer kids. I can remember being infuriated by my (then) 8 year old daughter’s school which refused to set any homework activities as “they only help middle class kids, so out of fairness we don’t give homework to any of the kids”. The levelling down stills sends a chill down my spine, but we could address the point by levelling up. Having a longer school day allows for homework to be both set and completed within school hours, like the “prep” in the poshest of schools. Let’s not forget that the better-off will help themselves to more education. There are the elite boarding schools. Then there are the private day schools with 8 hours a week more teaching, plus often Saturday lessons and / or sport. In the state sector, there is the booming industry in tutors. In Japan, for all its schooling, 45% of kids over 14 are spending an average of a further 5 hours per week at the Juku, getting private tutoring. 

It is often said that we should only make a move like this if we have evidence. Well, to a point. The current UK debate is missing even that point. There is a lot of flailing around comparing international countries, their education success and the amount of schooling they get. There has been a recent flurry of comparing UK school hours to those who do better than us in international educational tests. But it is impossible to isolate just this one variable when comparing UK or US educational outcomes with, say, South Korea on the one hand and Finland on the other. The best way to find evidence is to look at variation in hours within similar schools within similar areas in the same country. There is now growing evidence from the US that extended hours makes a big difference – and more difference than many of the other things into which lots of money is being poured ( class size, better teachers, etc). Let’s look at specific examples.The Expanded Learning Time experiments in Massachussets has added 2 hours per day to pilot schools. The effect? After just 1 year, there was a 44% boost in maths proficiency, 39% in English and 19% in Science. And the achievement gap narrowed too – by 35% in English. The 57 KIPP schools in the US achieve remarkable results in the US’ most deprived areas. With 80% of kids from low income families and 90% African American or Latino, the KIPP schools’ results are 2 to 3 times better than similar schools elsewhere. The school day at KIPP? It’s 7.30-5.00, with additional Saturday sessions and summer schools too. Some smart research in the US has proven the opposite effect when school hours are reduced – when schools are closed by snow, achievement falls, in proportion to the time closed. The evidence is getting louder. In 2013, 5 US States extended learning hours by 300 hours ( the equivalent of 10 weeks extra teaching). 

What about the teachers? In schools with extended hours, the biggest supporters have become the teachers. But surely they must resent the extra hours, the longer days? Aren’t they already at breaking point? How can they be supportive? Firstly, having more time each day, means that lessons are less rushed, less stressful, more relaxed. There is more time on the task – time to explain, to repeat, to explore. Secondly, schools with extended time find new ways to free up teachers from teaching. For example, the US elementary school where there is a big PE session everyday ( not just a small one once a week as before) taken by outside sports coaches and supervised by teaching assistants. During these sessions, teachers within the same grade get time to plan work together. Thirdly, some schools have added extra weeks onto the year without reducing teacher holidays by staggering teacher leave across the year. Fourthly, schools have brought a more diverse mix of volunteers, community groups and businesses into the longer curriculum, using the extra hours for an exponential increase in cultural experiences. Finally, schools with more time think harder about how to organise that time. Using data to individualise programmes for kids, they begin to catch up with the technological ability of their kids and let them access the almost infinite range of online activities. 

In simple terms what we are doing is giving kids the equivalent of an extra 7 years of compulsory education between the ages of 5 and 16 and giving teachers almost no time constraints. This is not about creating Gradgrind Academies which didactically stuff every waking hour with facts, figures and tests. It is the opposite. It’s about creating a lot of space in the day for play, creativity, relaxation, exploration and exercise. it’s about creating acres of space across the year for trips, volunteering, personal projects, work experiences, character building and bonding. It’s about making sure that kids don’t fall behind or fail to understand. It’s about schools opening their doors to let the wider community come in to help nurture and educate local kids. It’s about disrupting the current inertia so much that schools really do fundamentally re-think how their school operates. 

What about the money? It’s bound to need some new money. But we already have the buildings and the staff to cover most of it. There is existing money to recycle. An obvious example is teaching assistants and what they currently do. There is now roughly 1 teaching assistant for each 2 teachers. And yet the evidence shows that not only do they have no positive impact, but they seem to have a negative impact. Kids are better off without them. That clearly comes down to how they are used. One simple way to redeploy this resource in the 45/45 school year which I’m proposing is for TAs to supervise group activities independent of teachers. If the 200,000 or so TAs were in charge of activities (e.g. supervising homework or enrichment activities or online education sessions), this could leave teachers free to use large parts of the longer day for preparation and marking. We also need to tap into the volunteering spirit in local communities – what is stop us bringing the outside world into the school? For example, why not bring Scouts, Guides and Cadets into the school day? Or local football clubs? Or dance classes? The extra-curricular could come into the curriculum – but for all kids, not just those with supportive or able parents. We shouldn’t rush to put large amounts of money into the system. Necessity is definitely the right mother for the invention needed to take advantage of all this new time, whilst giving teachers the time they need beyond teaching. But new money should be found. More people working, less youth crime – the immediate fiscal benefits can pay for a good deal of any extra cost. But the real payback comes from a transformed generation who will leave their extended education with far higher skills and earning potential than the current generation. That will pay-off more than any infrastructure projects. So we should make this our top investment priority – creating a new category in capital expenditure rules for “human capital”. 

Is this idea too much too soon? Well, I always think a good way to test out an apparently too bold policy idea is to ask “If this new idea had been well established for the last 20 years and we proposed scrapping it, what would be the public reaction today? Relief, indifference, opposition?” Let’s assume that today’s parents had grown up expecting that schools were open 45 weeks a year and 45 hours per week. This fitted into their full time jobs. They got the same holidays as their kids and a working day that fitted inside school hours. Their kids had a broad and rich education, with lots of enrichment. And then, in order to please teachers and save a little money, the Government of the day proposed closing schools for 7 weeks a year and shortening the school day by 2.5 hours. Suddenly, a couple of million staff (mostly women, probably) would have to give up work, or go part-time. School-life would be pared down to the bone – a crammed day, with stressful lessons, kids falling behind, kids falling out, no time of the sports or arts, no place for the community in the curriculum. There would be uproar. An Election promise to go back to what we actually have today would be the biggest vote loser in history. So why wouldn’t an Election promise for my 45 / 45 model be the biggest vote winner since 1945? It must at least be good for a 45% share of the vote. 45 – remember the number!