THE MAIN forms of federal aid to higher education are up for renewal in this Congress. The bill extending them may be the most important social legislation of the next two years. The government has no programs more important to equal opportunity, nor any in a greater tangle. Nearly half the nation's 12 million college and other post-secondary students now receive federal aid at a cost of more than $10 billion a year. But not all the aid is rationally given. There's a lot the Congress could usefully do.

1. Most of the aid is distributed through two large programs -- grants for the poorest students plus partially subsidized guaranteed loans reaching up into the middle class. The loans are an entitlement -- the subsidy each year is automatic for all who qualify -- while the grants are subject to the annual appropriations process. That's the wrong way around. The automatic spending should be on the neediest students; if any is discretionary, it should be the spending on the better-off.

2. The loan program is now run through the nation's banks which, though they incur little risk, are paid sizable premiums to take part. The government could raise and lend the money cheaper itself. In the past, budget accounting rules militated against this. Direct loan programs counted the same as grants, interest subsidies counted less and mere guarantees (until defaults occurred) counted not at all. Last year's budget agreement changed this. Direct loan and loan guarantee programs will both be recorded in the budget in terms of their likely eventual cost, which will put them on a similar footing. This may finally permit the government to set up a direct student loan fund without the appearance of busting the budget (which it would not be), and such a proposal has already been heard within the administration.

3. College costs have risen faster than federal aid, and the mix of aid has changed so that a smaller share is now in the form of grants, a larger one in the form of loans (and future indebtedness). That has compounded the difficult distributional problems that have always dogged these programs -- whether and how to concentrate the aid on those in greatest need and what formulas to use to spread the aid between expensive private institutions and less expensive public ones. Should a student at Harvard have the same percentage of expenses paid as a student at a state university, or only the same amount? The alternatives have very different consequences.

4. No one wants the federal government to get into the business of cost and quality control in higher education. No one wants it to waste money, either. A lot of aid to higher education goes to students not in colleges and universities but in trade schools. An entire private trade school industry has grown up around the federal programs. It is responsible for a disproportionate share of the defaults that now consume more money (approaching $2 billion a year) than the interest subsidies on student loans. Should so much of the aid continue to go to trade schools? What standards should these schools -- businesses, really -- be held to? How can the government begin to make cost-benefit judgments about them and not about all the other institutions whose fees it is also helping to pay? To what extent, if any, should the government try to impose price controls on colleges in the same way it has tried to do on hospitals and doctors, a large share of whose bills it also pays?

If only by looking the other way -- another kind of default -- the administration and Congress will answer all these questions in the next two years. The public should play a role in the debate. It is too important to be left to those in and out of government for whom these programs are a living.