A loan repaid in full under this final option does indeed create a net return for government, but this does not imply a profit for government.  Loan programs the world over suffer huge losses from defaults, and without exception programs which charge these higher rates use the surplus to offset these defaults.  In this sense, this option provides from cross-subsidizing across the student body, with successful beneficiaries subsidizing those students unable to repay their loans.

Though these are the core four options for loans, there are some twists that can be added.  One twist is to use these four policies not as absolutes, but as figures to which actual policy can be pegged.  Malaysia, for instance, has in the past a policy of charging interest equal to “inflation minus one percent”; Sweden has a policy of “government rate of borrowing plus one percent”, etc.  Thus, the actual rates are linked to one of each of the four options without following it exactly.

Another twist is to apply different policies depending on whether the borrower is in school or in repayment.     For instance, the US and Canada charge nominal zero rates while students are in school, and higher rates afterwards (in the US, the rate differs among loan program but is pegged to the government rate of borrowing; in Canada it is linked to the Prime rate).  A third twist is to have different types of loans for different types of students.  Japan provides zero nominal interest loans to students with very good secondary school results and loans at prime to students with weaker results.  In the same vein, the US offers more expensive (“unsubsidized”) loans to wealthier students while providing subsidized ones to students from less affluent backgrounds.

A low-inflation world means loan subsidies are a lot cheaper to implement than they were, say, twenty years ago, but they are not costless.  And it’s very hard to argue that interest rate subsidies actually increase access.  There have been some substantial policy changes in loan rates across countries over the last couple of decades and no one has credibly come forward with evidence to suggest that these rates make any difference to application or enrolment rates.

For the most part, the economic effects of loan subsidies consist of increasing the purchasing power of educated mid-to-late 20-somethings.  If you think this is a group worth subsidising, then you should be in favour of student loan subsidies.  If not, you probably should want student loan subsidies to be kept to a minimum, and the money used for things which are actually proven to increase access (such as income-targeted grants).

That said, there’s policy and there is politics.  At the moment the pendulum in most of the world is to reduce interest on student loans – and certainly to avoid anything that looks like a market rate.  Fair enough: but that’s no reason to go overboard.  A Dutch solution – providing loans to students at government rate of borrowing for the life of the loan – is a good middle-ground solution.  Governments do not subsidize these loans, but students get a far-better-than market rate nonetheless.  A reasonable compromise all around.