3 ms·

Do we have any silver bullets? My understanding is most social services show very small effects once replicated.

TaylorAlexander 4y ago
I think this understanding is incorrect. My understanding is that housing services, mental health services, medical services, transportation infrastructure, and related things like maternity and paternity leave, good labor rights, good pay and fair working conditions all lead to a reduction in low level crime and homelessness. There is no need for a single silver bullet if we have a collection of policies that work. To just pick one issue here is a nice article about housing: https://charterforcompassion.org/shareable-community-ideas/public-housing-works-lessons-from-vienna-and-singapore https://charterforcompassion.org/shareable-community-ideas/p...
- JamesBarney 4y ago
  The fact that some places with affordable housing have much lower crime rates is incredibly far from a replicated RCT (or RCT like data).
  - TaylorAlexander 4y ago
    This is true, but I think RCT-like conditions are hard to create in the social sciences. This does not mean we are unable to make informed policy choices based on the available data.
    - JamesBarney 4y ago
      For interventions they're not that hard, and 99% of correlations that suggest interventions fail to find an effect when they try and RCT. While it's true it's evidence, it is very very very weak evidence.
      - TaylorAlexander 4y ago
        > 99% of correlations that suggest interventions fail to find an effect I'm genuinely trying to square your thinking with the specific services I listed in an earlier comment: > housing services It seems obvious and intuitive to me that providing discounted housing services would clearly have an effect. Clearly if we can manage spiraling housing costs, more people can stay housed and fewer people will be destitute. And the case in Vienna supports this notion: https://www.huduser.gov/portal/pdredge/pdr_edge_featd_article_011314.html https://www.huduser.gov/portal/pdredge/pdr_edge_featd_articl... Have you found an RCT-like study that provided affordable housing of this quality and failed to find an effect? > mental health services I would be surprised if you found that providing quality mental health services does not have an effect. > medical services This is another one which seems obvious and intuitive to me. A fair number of people in the USA end up homeless or destitute due to medical debt or lack of appropriate medical care. Every other major country offers some form of free medical care and I cannot see how this would fail to have an effect. > transportation infrastructure If you follow the urbanists on youtube or websites like Strong Towns it seems clear that requiring everyone to have a car places a heavy burden on the poor, and it seems obvious to me that if you have good transportation infrastructure (high speed rail between cities and good metro rail in cities, appropriate zoning for walkable neighborhoods), then it becomes much easier for the poor. This is not something that you can easily study as an RCT. > and related things like maternity and paternity leave, good labor rights, good pay and fair working conditions Again these are obviously beneficial. I feel like I provided a lot of good examples and I am surprised by the continued assertion of "very very very weak evidence". Really? For every one of these things?
        JamesBarney 4y ago
        Let me start with what I agree with, we should try all of them in a fashion that tests if they work, and if any of them did role them out everywhere. I also agree that many of these have beneficial effects besides a reduction in crime and we should maybe provide these services anyway because it makes peoples lives better. I also agree we should manage spiraling housing costs by upzoning everywhere. I think affordable housing could actually make things worse by concentrating poverty. I'm not convince transportation infrastructure helps the poor more than direct cash transfers. The US is really terrible at building cost effective transportation infrastructure, and most of the places that have it are terrible places to be poor like San Francisco or New York. The only example I need is one replicated natural experiment or RCT that shows any of these working.

I think this understanding is incorrect. My understanding is that housing services, mental health services, medical services, transportation infrastructure, and related things like maternity and paternity leave, good labor rights, good pay and fair working conditions all lead to a reduction in low level crime and homelessness. There is no need for a single silver bullet if we have a collection of policies that work. To just pick one issue here is a nice article about housing: https://charterforcompassion.org/shareable-community-ideas/public-housing-works-lessons-from-vienna-and-singapore https://charterforcompassion.org/shareable-community-ideas/p...

JamesBarney 4y ago
The fact that some places with affordable housing have much lower crime rates is incredibly far from a replicated RCT (or RCT like data).
- TaylorAlexander 4y ago
  This is true, but I think RCT-like conditions are hard to create in the social sciences. This does not mean we are unable to make informed policy choices based on the available data.
  - JamesBarney 4y ago
    For interventions they're not that hard, and 99% of correlations that suggest interventions fail to find an effect when they try and RCT. While it's true it's evidence, it is very very very weak evidence.
    - TaylorAlexander 4y ago
      > 99% of correlations that suggest interventions fail to find an effect I'm genuinely trying to square your thinking with the specific services I listed in an earlier comment: > housing services It seems obvious and intuitive to me that providing discounted housing services would clearly have an effect. Clearly if we can manage spiraling housing costs, more people can stay housed and fewer people will be destitute. And the case in Vienna supports this notion: https://www.huduser.gov/portal/pdredge/pdr_edge_featd_article_011314.html https://www.huduser.gov/portal/pdredge/pdr_edge_featd_articl... Have you found an RCT-like study that provided affordable housing of this quality and failed to find an effect? > mental health services I would be surprised if you found that providing quality mental health services does not have an effect. > medical services This is another one which seems obvious and intuitive to me. A fair number of people in the USA end up homeless or destitute due to medical debt or lack of appropriate medical care. Every other major country offers some form of free medical care and I cannot see how this would fail to have an effect. > transportation infrastructure If you follow the urbanists on youtube or websites like Strong Towns it seems clear that requiring everyone to have a car places a heavy burden on the poor, and it seems obvious to me that if you have good transportation infrastructure (high speed rail between cities and good metro rail in cities, appropriate zoning for walkable neighborhoods), then it becomes much easier for the poor. This is not something that you can easily study as an RCT. > and related things like maternity and paternity leave, good labor rights, good pay and fair working conditions Again these are obviously beneficial. I feel like I provided a lot of good examples and I am surprised by the continued assertion of "very very very weak evidence". Really? For every one of these things?
      - JamesBarney 4y ago
        Let me start with what I agree with, we should try all of them in a fashion that tests if they work, and if any of them did role them out everywhere. I also agree that many of these have beneficial effects besides a reduction in crime and we should maybe provide these services anyway because it makes peoples lives better. I also agree we should manage spiraling housing costs by upzoning everywhere. I think affordable housing could actually make things worse by concentrating poverty. I'm not convince transportation infrastructure helps the poor more than direct cash transfers. The US is really terrible at building cost effective transportation infrastructure, and most of the places that have it are terrible places to be poor like San Francisco or New York. The only example I need is one replicated natural experiment or RCT that shows any of these working.