Prisoner release with accountability
Idea: How to induce sorting and accountability for prisoners.
El Salvador accomplished an amazing feat, they went from one of the highest murder rates in the world to one of the lowest in a couple years. This was accomplished by a massive incarceration campaign against all of the gangs and was done without any due process. This wildly successful and transformative move has brought peace and hope to a small country that was suffering under violence and corruption. The problem they are facing now is that they locked up 70,000 people (astonishingly only losing 8 police in the process) and those locked up did not have a trial or a sentence. I was wondering what would be the best way to sort a prisoner population so that you freed those members who would become productive members of society while not letting lose the murderers.
The current system would most likely employ a large group of lawyers, psychologists and social workers and task them with attempting to understand each individual and then make a judgement call about each one. If they were correct, then things work out fine but if they are wrong then there is finger pointing and blame, but no real accountability. In other words they are not incentivized to be correct and are actually incentivized to keep the status quo as they face a possible downside risk with a poor choice versus virtually no upside if they parole someone. I propose a system of public prisoner voting combined with accountability.
The prisoners themselves are the ones who know this population the best, and so they would be best at self sorting, but how could they be trusted to do this in a way that the society they would be returning to would be pleased with. They should be given a vote. The vote would be completely public and individuals couldn’t vote for themselves. They would be voting for who to release. The prisoners with the most votes would be paroled. If they commit a crime during their parole then they would be locked up again and all those that voted for this individual would lose their eligibility to be up for parole
.
Hypothetical example to judge if the incentives are aligned
First, the state could determine the eligible members of this group, for instance maybe anyone already had a record may not be in the first group or maybe there is enough known about who the gang leaders are and so they wouldn’t be included. Once we have a population to start with and it is decided that this program will parole 100 prisoners. The prisoners are given the vote. What if the gang leaders (highest ones eligible) all decide they want to be released and they are the most popular amongst the prisoners, because of fear, loyalty, bribery, etc.? Well if this was the case and they immediately went out into society and returned to crime and were caught doing so then they get locked up and all of the prisoners who voted for them lose their chance. The problem the leaders would face if they decided on this path is that they would either have to do the crime themselves or recruit new members (or members who hadn’t been caught). They would be monitored closely the entire time. So say that leadership organized the majority of the voting and got 90 members out of the 100.
If all 90 returned to crime and were caught then they would all be returned to prison and all those who voted for them would lose their option to be on the next ballot. The other 10 that did not return to crime would then have all those who voted for them as part of the next possible group of candidates for selection. This aligns the incentives for the people who are voting. The highest goal of a prisoner obtain as many votes as possible so that they would be selected to be paroled, but who you cast your vote for would be very important if you were not going to be one of the winners. You would be incentivized to select the individual who best satisfies only these 2 criteria: they would win enough votes to be selected and 2 that once on parole they would not behave in any way that would have parole revoked.
So when trying to decide who to vote for every prisoner would have to be making the case for why they were the most trustworthy person to behave properly once released. Only the prisoners who were able to convince enough people of this would be selected. The key difference between this and normal voting is how accountable the action is to the voter (more to say on this another day). If I vote for someone who does a objectively terrible job, starts wars and gets people killed and spends tons of money, well those costs are spread equally between the majority who voted for and the minority against. In the prisoners case, the best chance I have of obtaining parole would be being able to identify the person who would actually behave on parole and could convince enough people they would. A prisoner who was great at convincing others they would behave and then doesn’t would be returned to a prison with all the people who voted for them, they would suffer the consequences of their actions.
What if all 100 of the first group was bad and were all returned to prison, well then a large portion of the sorting has been accomplished, now you have the 100 and all the people who voted for them ineligible for parole. This next ballot would consist of far fewer candidates and the incentives would align to now better select the best people. The state would always reserve the right to limit whomever was on the ballot or adjust the basic parameters.
Making accountability direct drives systems towards serving the incentives that exist. This seems like a good way to balance the needs of a society while incentivizing the prosocial behavior of prisoners and efficiently sorting in the process. I think it could be useful in El Salvador but the US also has a lot of prisoners who could be productive members of society if given the right incentives and opportunities.