I agree, but the situation is even worse than that. The Rick Perrys of the world think that convictions of any type, including false convictions, are Just Fine-- they don't even see that there's a tradeoff to be made.
In signal detection theory, there's a distinction made between "bias" (which is what you're talking about, "which way to skew") and "sensitivity" (also called "discriminability"), which is, at heart, how good the "test" (here, the justice system) is at distinguishing between "signal" and "noise" in the first place.
In theory, everyone, no matter what their bias, should be in favor of improvements to the justice system that provide better discriminability (assuming they're not too expensive, etc.). After all, discriminability improvements lessen both false negatives and false positives (or, at least, decrease one while leaving the other unchanged). And yet, in reality, many "tough on crime" types oppose such improvements.
For example, DNA testing is a scientific advance that provides a wonderful increase in discriminability, in those cases where physical evidence exists. Use of DNA testing should be pretty much automatic, not only in new cases but to reopen old ones. There's essentially no way you're going to let an already-convicted, actual guilty person go free in these cases, but you might free some innocents. However, many judges have ruled against DNA testing, even though the defense is willing to pay the expense themselves. Even worse, in many (most?) jurisdictions, as a matter of routine, physical evidence destroyed after a trial is concluded.
There are a few "good guy" prosecutors like Craig Watkins of Dallas County, who have actively sought to improve the discriminability of the justice system. However, the vast majority of prosecutors seem to feel that "more convictions = good" and that false positives (convicting innocents) are Just Fine-- i.e., it's not merely that they want to "skew" the bias such that little weight is given to false convictions, it's that they give no weight at ALL to them. They don't even see that there's a tradeoff to be made.
(Every year, for the "Methods" unit test in my psychology class, I ask a variant of the same signal detection question. :-) Maybe one of my students will someday be a better prosecutor.)
SDT
In signal detection theory, there's a distinction made between "bias" (which is what you're talking about, "which way to skew") and "sensitivity" (also called "discriminability"), which is, at heart, how good the "test" (here, the justice system) is at distinguishing between "signal" and "noise" in the first place.
In theory, everyone, no matter what their bias, should be in favor of improvements to the justice system that provide better discriminability (assuming they're not too expensive, etc.). After all, discriminability improvements lessen both false negatives and false positives (or, at least, decrease one while leaving the other unchanged). And yet, in reality, many "tough on crime" types oppose such improvements.
For example, DNA testing is a scientific advance that provides a wonderful increase in discriminability, in those cases where physical evidence exists. Use of DNA testing should be pretty much automatic, not only in new cases but to reopen old ones. There's essentially no way you're going to let an already-convicted, actual guilty person go free in these cases, but you might free some innocents. However, many judges have ruled against DNA testing, even though the defense is willing to pay the expense themselves. Even worse, in many (most?) jurisdictions, as a matter of routine, physical evidence destroyed after a trial is concluded.
There are a few "good guy" prosecutors like Craig Watkins of Dallas County, who have actively sought to improve the discriminability of the justice system. However, the vast majority of prosecutors seem to feel that "more convictions = good" and that false positives (convicting innocents) are Just Fine-- i.e., it's not merely that they want to "skew" the bias such that little weight is given to false convictions, it's that they give no weight at ALL to them. They don't even see that there's a tradeoff to be made.
(Every year, for the "Methods" unit test in my psychology class, I ask a variant of the same signal detection question. :-) Maybe one of my students will someday be a better prosecutor.)