I see economic theory applied to the doping situation.
Economic theory says the vast majority of the people is risk averse. The gain of doping should be way bigger than the possibility of getting caught. It is often that riders only dope when they think they found a new substance that is not recognizable in tests (think of the whole CERA episode with Ricco, Schumacher, Kohl, Di Luca et al). All those riders had a significant increase in performance in that very year. All were caught.
Which means CERA is very effective (which it is), but also other doping is not as effective, following the line that everyone is doping (or at least the 30 best). With CERA we saw a sudden jump forwards by certain riders. Looking at the best riders of today I can only conclude that their progress was very gradual and predictable (with, of course, some exceptions). Which means they must be using the not-so-effective dope for years now, or they must be clean. If the best 30 are doping for years, they will know that the day one of them is caught with the not-so-effective doping will come closer every day. Following the risk aversion theory, they must at least try to ride without the not-so-effective doping for once. When they notice that the doping is not-so-effective (which it isn't, if we look at the gaps which were way bigger in the proven doping years, and looking at the progress young riders can make compared to that years), they will not dope instead, as they are risk averse.