Statistics show us that colleges have been spending more and more per student as time passes. As to the reason for it, there have been a lot of theories and debate over it. I'd like to shed some light on two major conflicting theories by economists here.
The first is Baumol cost disease. Originated by William Baumol, a highly influential economist, he argues that increased productivity demand increased wages but also increases wages in competing industries that don't see increased productivity. In our educational setting, this means that schools are having to compete with the private sector over talented professors. This means they have to pay salaries beyond what they would normally do without the university being any more productive. And this increased spending is in turn causing increases in costs for students. A far more detailed and sophisticated version of Baumol hypothesis is advanced in this book (http://www.amazon.com/Why-Does-College-Cost-Much/dp/0199744505) by economists Robert Archibald and David Feldman.
But looking at actual data from the Delta Cost Project, numbers don't seem to overly support the theory. While expenditures on "instruction" have gone up over the years, it doesn't seem to be commensurate with increases in overall costs. There have been greater increases in other areas so Baumol cost disease doesn't seem to be the core of the problem.
The next theory is the Bowen hypothesis which argues quite the opposite. Elaborated in the book "Costs of Higher Education" by economist Howard Bowen, the basic idea is that most universities being non-profits spend all the money they raise whether they need it or not. This means that they end up spending a lot on recreational facilities, fancier buildings etc. which don't really improve the quality of education. And they try to raise as much money as possible in the first place.
This means colleges could spend much lesser and still maintain the quality of education. They're not raising tuition because they have to but because they want to. And one main reason this is possible is because higher education is an experience good - a commodity that is unknown before purchase. According to economists experience goods markets can be efficient when it is purchased frequently because you learn from the first time if it is worth it. But unfortunately a college degree is a one time deal and there is no need for satisfying repeat customers. And information regarding quality of colleges and a student's fit within can be extremely hard to obtain.
One major perceived indicator of quality by students in the sticker price. A more expensive school must mean that the quality of education is better. And this is a great incentive for schools to hike up prices. Worse still, schools that try to compete on the basis on cost are viewed by the consumers as cutting quality. This leads to no cost competition and hence no pressure to reduce costs at all. And by Bowen hypothesis, this in turn leads to more spending and the vicious cycle continues.
The data seems to back Bowen hypothesis better. Expenditures on "institutional support" increases were higher than "instruction" at most universities. While both theories have their supporters, Bowen hypothesis seems to be favored with the recent data coming in.