> If you children are getting killed, beaten, and imprisoned by the legal system, and have no access to good education, you might want to move a little quicker.
"People are suffering now" is not a good blanket argument for rapid change. First, it needs to be shown that the changes proposed will not be worse than the original condition.
To use your example, last summer many people were arguing for quickly defunding the police. While that would certainly solve the problem of children suffering under the legal system, it might also result in vigilante groups stepping up to enforce the law, or in increased urban gang activity, or in increased domestic violence, any of which might end up with increased overall victimization of children.
My personal opinion is that individual jurisdictions should experiment more at a small scale. Maybe SF abolishes the police, and Minneapolis increases funding for social work, and the Chicago police install robocops at every corner, and Nashville provides benefits only to families which stay intact. Then after two years it will be obvious which policies _really_ don't work, and after five years there will be a good handle on the set of policies that might work, and after fourty years most of the magnitudes of the possible second- and third-order effects will be understood, at least as well as it is possible to know them.
But it is impossible to have local experiments when the popular call is for immediate action on the national level. Alas, one is left with digging through the archves of history and trying to see where natural experiments have been done, and what confounding variables there were.