The HELP student loans scheme involves significant expense to government. One of these is the cost of carrying around $20 billion in accumulated debts while only getting CPI indexation. Figures DEEWR gave me put this cost at $650 million for 2009-10.
The discount for paying upfront and the bonus for voluntarily repaying early are designed in part to help minimise this cost (their other purpose is to avoid the risk of non-repayment).
The discount especially has sometimes been criticised as an ‘unfair’ lower price to ‘rich’ people who can afford to pay upfront, as one commenter on another post did. While we can sensibly debate the level of the discount, which is now 20% but has been 25% and 15% in the past, the idea of the discount is sound. Given the cost of the interest subsidy, it can be cheaper to write off the discount now than to carry the debt. This either saves money for taxpayers or makes more money available for other government purposes.
Using the same male arts and law graduates in professional jobs on median earnings examples in an earlier post, I compared the cost to government of the different payment and repayment options. I assumed that the government borrows money at 6% and lends at 2.5%.
Interest subsidy on student debt
As can be seen, in these examples it is cheaper for the government to give the discount than to lend the full amount, though only by a small amount in the case of full up-front payment. For the government, the optimal option in my scenario is the partial up-front payment. For these relatively financially successful graduates the interest cost on a reduced loan that is repaid fairly quickly is lower than the bonus amount.
The worst scenario for the government is the voluntary repayment at the end of the loan. This and borrowing the full tuition charge and paying back entirely through the tax system are the most expensive options because of the high amounts of lending accrued early on. The voluntary repayment makes things worse because the 10% bonus is more expensive than the interest on the dwindling sums owed towards the end of the loan.
In these examples the bonus is perhaps bit higher than it needs to be, but these are probably not typical examples. I need to work on more typical repayment trajectories, taking into account that most graduates are women who are less likely than men to work full-time.
Andrew Norton 
I hope that you take into account the fact that if a person is unlikely to work full time when they graduate, their rich parents will probably not pay back the HECS debt anyway.
Just a thought.
LikeLike
Lomlate – I am going to do what research I can, but I don’t think there is much on the characteristics of those who pay upfront or in the case of parents why they pay or how well they can predict the future earnings of their offspring.
But clearly selection effects are a factor in assessing payment options.
LikeLike
lomlate — I think that’s the idea of HECS. The government takes a hit for those that are unfortunate. I realize that there are categories that some people don’t like in this equation and wouldn’t be considered unfortunate for one reason or another (e.g., people that become full time parents etc.), although trying to specify these would be a minefield. An alternate to up-front fees if people are worried about how much the government loses would be to charge a higher level of HECS and use that money to subsidize the people that don’t pay it back, in which case there may be more arguments about this.
LikeLike
If the justification for HELP student loans is to overcome a financial markets failure and not provide a further subsidy, as you suggest in your previous post, then I think the income contingent aspect of the loans needs more attention than the low interest rate. Your analysis shows that with the various discounts and bonuses, the effective interest rate on HELP debt is higher than CPI. I think it also follows that the quicker one’s pay increases and one pays off the debt, the closer the effective interest rate on a HELP loan is to something commercial. So the real cost to the taxpayer from the loans arises from the tendency of graduates to defer repayment by travelling, undertaking further study or, most significantly, taking time off to care for children fairly soon after graduation. This is an extremely regressive policy. Who are the women who start having children within 5 to 8 years of graduating? I suggest they would be women in relationships with professional or high-earning partners. How beneficial is it to the next generation to have tertiary educated mothers? Beneficial enough to effectively give these women free education from working class taxes? How about giving one year’s grace following graduation and then requiring a fixed sum or proportion of the debt to be repaid back every year after, come hail or shine? I think $5k pa would be a reasonable sum, with no discounts and the debt indexed at the government bond rate. It would still be a middle-class subsidy, but a much less egregious one than at present.
If the government wants to encourage people to become teachers or nurses, it can do this more directly.
LikeLike
Rajat – The argument I had in mind was less fiscally (and politically) ambitious than that, in asking how could HELP be made cheaper to taxpayers without undermining the income-contingent aspect of the scheme. The US has something similar to what you suggest, with quite high rates of loan default (esp. during their current economic difficulties). The idea behind income-contingent loans was to control the economic risk involved for students.
LikeLike
Yes, I understand that many US student loans (federal loans in particular) are full recourse and even bankruptcy does not expunge them. Perhaps one solution is to require graduates to be in the labour force and registered as unemployed before they get a repayment reprieve. Graduates who are voluntarily out of the labour force can to required to keep paying.
LikeLike
I remember seeing this question in my old Business Finance textbook a while back.
LikeLike