- It is basically our entire Quant 3 course condensed into one paper; and
- It is about a very real child welfare issue (that my division does a lot of work with)
It is the 2nd of these I want to spend a few minutes on. The issue, greatly simplified is this:
Research indicates that borderline children put in out-of-home placements(ie foster care), on average, have worse long term outcomes than those who remain in home environments that are abusive or otherwise unhealthy. This paper concludes:
(There are) better outcomes when children on the margin (of getting an out-of-home placement) remain at home
So as a policymaker, what do you do with this? I think you continue to put kids in out-of-home placements, but I'm not at all sure. The reason for my opinion is because it seems to me that policy is faced with 2 propositions:
- Remove the kid from home and there is a 99% chance he/she lives to see adulthood, but a 50% chance they'll be involved with the criminal justice system/homeless/subtance abusers
- Keep the kid in their home and there's a 30% chance of bad outcomes, but only an 85% they survive childhood.
Now obviously my numbers are total crap, but you get the point. To me this gets to the ultimate question of: what is the government's primary responsibility to at risk children? Is it to ensure they live to see adulthood, or is it to maximize their chances to have a productive adulthood?
In this case, at least, those policy outcomes may be mutually exclusive.
2 comments:
The paper may condense all of quant three, but does it speak to Ethics in Policy at all (I haven't read it)? It sounds like you've boiled the question down to its fundamental issue: what is the role of government?
Taking a purely positive approach (ethics aside) you can compute the optimal policy. Using your probabilities (which I assume are contrived) and applying Bayes Rule (oh, Bayes Rule, you glorious thing) you would have to value a life of crime at about 40% of a normal life (41.6%, to be exact) to be indifferent between the two outcomes. So, if you think a life of crime is worth less than 40% of a regular Joe, on average, then the optimal policy is to leave the kids in their original homes.
Given that your probabilities are actually quite extreme, this number is likely significantly larger in reality. As XX% goes up it gets harder to justify taking kids from their homes.
There are probably actual value-of-life statistics for criminals in the literature. My guess is they would be something on the order of 60%-70% which makes it very difficult to argue for removal from biological families.
Of course, an ideal policy would use this type of analysis more deliberately--focusing only on the kids in the most dangerous households.
Post a Comment