Governments, like individuals, often like to believe their varying instincts and aspirations all fit comfortably together even when they don't. They prefer to try to keep these tensions under wraps and sometimes don't even like to admit them in private to themselves. And the coalition is a case in point.

One of its favourite claims is that, despite the fact that all sorts of welfare support is being removed from families on middle incomes, when it comes to the very poorest they are doing more than their predecessors. The pupil premium usually gets a mention here, followed by the expansion in student support for the most disadvantaged.

Another cherished claim is that punitive marginal tax rates for those struggling on modest incomes seeking to earn their way up will be reduced - a point  made with great passion by David Cameron in his 2009 Conservative party conference speech when he railed against an example of a 96 per cent tax rate hitting a single mother. At this point, the coalition tends to highlight the Olympian ambition of the Universal Credit and its effort to integrate benefits and tax credits and create a single, smoother means-test in our welfare system. 

A final claim is that it is right and proper to localise decision making about how to allocate scarce financial support – for instance through the decentralisation of council tax benefit.

Many would want to challenge each of these claims; but let’s leave that to one-side. What should be beyond dispute – though this is rarely recognised – is that these three agendas don’t make for comfortable bed fellows; indeed, they don’t really belong in the same room at all. Try and defend them individually if you so wish, but don’t pretend they add up to a coherent strategy.

The recent Child Benefit saga gave daylight to some of these tensions, demonstrating in vivid terms how poorly designed middle-class welfare retrenchment can generate nasty means-testing problems that then have to be mitigated.  As of next January the removal of Child Benefit from households with someone on over £50k will mean a new 50p or higher effective tax rate for these families if they have one child, and 60p for those with two kids. If the Budget hadn’t had so many other highlights the dragging of the 50p tax rate from the super-rich down the income scale would surely have received more notice.  

Now a new report by professor John Hills, perhaps the UK’s foremost authority on the welfare state (together with his LSE colleague Ben Richards), provides us with another dramatic case study of how different policy objectives combine to form a nasty cocktail. The perhaps unintended and unforeseen effect of a shift to greater private contribution in welfare (this time in the form of higher tuition fees), combined with efforts to protect the position of the very poorest (increased bursaries and grants aimed particularly at families with earnings under £17k), and a nod towards localism (universities run their own support system) is to create a new aspiration trap – truly eye-watering effective tax rates hitting families in low-to-middle income Britain sending a child to university this autumn.

This stems from the way in which the complex patchwork of student support gets withdrawn as household earnings rise. Some of the resulting ‘cliff edges’ soar high above those that triggered the Child Benefit row. 

To understand how this will actually play out in practice Professor Hills considers two families each with a child about to go to the University of Oxford. One family has earnings of £17k and the other £44k – so a difference in gross earnings of £27k.  After we take account of the impact of the overall tax and benefit system the difference between these families falls to £13,250. But once we factor in the additional impact of all the different elements of the new student support system the gap collapses to a grand total of £200 (yes, you read that right).

To restate: an initial difference in gross earnings of £27k between (broadly speaking) a low-income and middle-income family is completely wiped out. The withdrawal of student support, together with the tax and benefit system, creates an effective 99 per cent tax rate on earnings between £17k and £44k. There is no point being better off. And that’s before we consider some truly scary effective tax rates at particular points in the earnings distribution (see chart). I think it is fair to say that the coalition hasn’t fully got its head around the politics of this.

And don’t think this is just some quirky Oxford phenomenon, though the issue is most dramatic there (which, to be fair, is because the support on offer at Oxford for those with the very lowest incomes is most generous). Hills surveyed the support at our largest 52 universities representing 60% of all HE students and finds that it is common for small differences in parental earnings to lead to several thousand pounds less support. 

Now, we can ask questions about aspects of this. There will be some parents who couldn’t care less about the financial position of their 18 year old, so they won't view a drop in support as any sort of hit on the family budget. And the report significantly understates the extent to which the withdrawal of some financial support, like cash bursaries and maintenance grants, may feel more like an effective 'tax rise' on the family than others, such as the removal of fee discounts (which may seem like a problem for the student tomorrow rather than the family today). Despite this the overall argument is strong.

The usually understated Hills concludes that despite the towering rhetoric about the what the universal credit will achieve, some parts of government are moving in “precisely the opposite direction”,  giving rise to new poverty traps. “It looks as if we will see a lot more of this in the future.  Already councils have each been told to work out their own way of making savings on Council Tax Benefit, which could result in them withdrawing benefit faster, adding to the poverty trap – but with rates and rules varying across local authority boundaries.  With budgets under pressure, it’s an obvious reaction to withdraw services from those with higher incomes, while keeping means-tested support for the poor. But what may seem a reasonable response to fiscal constraints while trying to protect the poorest in one sphere may overlap chaotically with other attempts to do the same thing”.

If this new twist to the student finance reforms sparks to life, as its impact on household budgets becomes clearer, it will send a shiver down the spine of leading members of the coalition. At the time they took the decision to bring in the new funding system they knew, of course, they were in for some choppy politics and that those on middle and high incomes would take a hit. But I very much doubt they grasped that it would lead to what many of them would consider to be totalitarian tax rates being imposed on families in the so-called striving classes whose kids are aiming high.