You could improve the situation by incentivizing people to identify cheaters and prove their cheating. If being a successful cheater-hunter was a good career, the field would become self-policing.

This approach opens its own can of worms (you don't want to overdo it and create a paranoid police-state-like structure), but so far, we have way too little self-policing in science, and the first attempts (like Data Colada) are very controversial among their peers.