> Well, the change that prevents this from happening again is DNA evidence.
This is the very optimistic interpretation - another tool and a permanent reduction in the wrongful conviction rate. The pessimistic interpretation is that DNA tests served as a short-term oracle where the system nigh-magically was stripped naked and its true past error rate laid bare, but to which the system will slowly adapt and figure out how to circumvent.
We can probably expect in a few decades to start seeing articles about how DNA evidence gets lost, misinterpreted, planted (remember OJ and the issues with the blood evidence? more than a few police have no issues with framing a guilty man, it seems), and other such strategies that are not obvious to people outside the system.
In other words, if there were suddenly another nigh-magical breakthrough which gave us a second look at true error rates, we would see, superimposed on whatever other trends are going on, a dip in the '80s-'90s and then a slow increase to whenever the second breakthrough's innocence projects got going.
Unfortunately, this is a very long-term prediction and there may never again be a breakthrough as dramatic as DNA testing which would give us the second oracle, so who knows if time will let us distinguish between the optimistic and pessimistic interpretations.
This is the very optimistic interpretation - another tool and a permanent reduction in the wrongful conviction rate. The pessimistic interpretation is that DNA tests served as a short-term oracle where the system nigh-magically was stripped naked and its true past error rate laid bare, but to which the system will slowly adapt and figure out how to circumvent.
We can probably expect in a few decades to start seeing articles about how DNA evidence gets lost, misinterpreted, planted (remember OJ and the issues with the blood evidence? more than a few police have no issues with framing a guilty man, it seems), and other such strategies that are not obvious to people outside the system.
In other words, if there were suddenly another nigh-magical breakthrough which gave us a second look at true error rates, we would see, superimposed on whatever other trends are going on, a dip in the '80s-'90s and then a slow increase to whenever the second breakthrough's innocence projects got going.
Unfortunately, this is a very long-term prediction and there may never again be a breakthrough as dramatic as DNA testing which would give us the second oracle, so who knows if time will let us distinguish between the optimistic and pessimistic interpretations.
(Related ideas: http://en.wikipedia.org/wiki/Risk_homeostasis & http://en.wikipedia.org/wiki/Goodhart%27s_law)