William Raspberry {op-ed, Oct. 7} says that Education Secretary William Bennett's new program of income contingent loans (ICLs) for college students is an "admirable" idea based on a sound principle: that students are the prime beneficiaries of their education and so should bear at least some of the cost. I agree. But Mr. Raspberry also has reservations about ICLs, derived from an analysis by Clemson University Prof. Robert J. Staaf that is badly flawed and highly inaccurate.

Prof. Staaf makes sweeping claims about ICL's higher default rates, but offers no evidence whatsoever to support his contention. In fact, our experience with why and when students default suggests the opposite is true. Prof. Staaf also predicts that students will avoid unsubsidized ICLs in favor of subsidized guaranteed student loans. This ignores the fact that 10 colleges and universities are making ICLs right now, and they are having no trouble finding plenty of takers. Let's remember too that the administration originally proposed a full-scale ICL program, not the pilot eventually agreed to by Congress, and also called for reducing subsidies in the GSL program. Under these circumstances, income contingent loans would be very attractive indeed.

We recognize that ICLs aren't perfect; that's why we're recommending a number of improvements, including expanding eligibility to graduate students. But we shouldn't throw out the baby with the bathwater. The ICL concept is, as Mr. Raspberry acknowledges, a good one, and ICLs should play an important role in financing college costs for years to come. BRUCE CARNES Deputy Undersecretary U.S. Department of Education Washington