Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Investigating a ten-year-old estimate that “most social programmes don’t work”

80000hours.org

Post reply on HN
Print view View original

11–20 of 86 posts

Re: Investigating a ten-year-old estimate that “most social programmes don’t work”

#11

Post by confounded »

nonbel wrote: post #10

The way of telling if it works is: statistical significance. This "approach" to research was debunked as pseudoscience long ago (eg here is one from 1967[1], and that wasn't the first)... So the answer is: no one knows what works or doesn't, (edit: at least not based on the usual claimed method of drawing conclusions from the data). [1] http://www.fisme.science.uu.nl/staff/christianb/downloads/me...

This is a rather broad debunking of the p-value approach to establishing scientific knowledge, and not specific to social programs/programmes.

Re: Investigating a ten-year-old estimate that “most social programmes don’t work”

#12

Post by fatalogic »

As a former social worker I think most social programs don't work because they try to deal with multifaceted problems from a single point of action. It's like trying to fix a suspension bridge by replacing one cable while all the other cables are broken.

Re: Investigating a ten-year-old estimate that “most social programmes don’t work”

#13

Post by nonbel »

confounded wrote: post #11
nonbel wrote: post #10

The way of telling if it works is: statistical significance. This "approach" to research was debunked as pseudoscience long ago (eg here is one from 1967[1], and that wasn't the first)... So the answer is: no one knows what works or doesn't, (edit: at least not based on the usual claimed method of drawing conclusions from the data). [1] http://www.fisme.science.uu.nl/staff/christianb/downloads/me...

This is a rather broad debunking of the p-value approach to establishing scientific knowledge, and not specific to social programs/programmes.

Yes, I first came to this conclusion with regards to healthcare (which is where I have seen the sausage made).

Re: Investigating a ten-year-old estimate that “most social programmes don’t work”

#15

Post by confounded »

robertwiblin wrote: post #2

I'm one of the authors of this post - look forward to hearing you comments!

It would be good to get your read on whether this is absence of evidence, or evidence of absence. To me (I work in stats) it sounds like the former, but the general reader could get the impression of the latter, especially given the title. What's your take?

Re: Investigating a ten-year-old estimate that “most social programmes don’t work”

#16

Post by cuchoi »

nonbel wrote: post #10

The way of telling if it works is: statistical significance. This "approach" to research was debunked as pseudoscience long ago (eg here is one from 1967[1], and that wasn't the first)... So the answer is: no one knows what works or doesn't, (edit: at least not based on the usual claimed method of drawing conclusions from the data). [1] http://www.fisme.science.uu.nl/staff/christianb/downloads/me...

It might not be perfect, but it is the best way we have come across.

Re: Investigating a ten-year-old estimate that “most social programmes don’t work”

#17

Post by jake-low »

robertwiblin wrote: post #2

I'm one of the authors of this post - look forward to hearing you comments!

Thank you for including such a thorough and approachable conclusion. As someone with no background in statistics, it helped me understand the content.

Genetic mutations seem to have similar properties (by analogy). Most don't 'work', but evolution still uses random mutation to build better organisms. Progress is sensitive to the degree of selection pressure: too small, and unfit organisms will reproduce; too great, and even fit organisms will frequently die.

I wonder if there's an empirical way to determine the optimal degree of selection pressure (where "optimal" means the degree which most rapidly improves the organisms) and whether the same logic could be applied to determine what fraction of underperforming social programs to cut and replace with new programs.

Re: Investigating a ten-year-old estimate that “most social programmes don’t work”

#18

Post by nonbel »

cuchoi wrote: post #16
nonbel wrote: post #10

The way of telling if it works is: statistical significance. This "approach" to research was debunked as pseudoscience long ago (eg here is one from 1967[1], and that wasn't the first)... So the answer is: no one knows what works or doesn't, (edit: at least not based on the usual claimed method of drawing conclusions from the data). [1] http://www.fisme.science.uu.nl/staff/christianb/downloads/me...

It might not be perfect, but it is the best way we have come across.

What about actual science? Where you figure out how to get reproducible results, come up with theories to explain why those results are the way they are, then make predictions with the theories to test on new data?

Re: Investigating a ten-year-old estimate that “most social programmes don’t work”

#19

Post by BartSaM »

I would love to see more work-for-benefits programs. Where you are required to do some labour work for the community (cleaning wild areas, swiping side walks, planting trees etc.) in exchange for some better than average benefits. IMO everyone should be entitled to such a work. This way we would eliminate involuntary extreme poverty.

Re: Investigating a ten-year-old estimate that “most social programmes don’t work”

#20

Post by schneems »

My S/O used to work at a company AuntBertha. They're a "search engine for social services" but they are a for-profit company (a B-Corp).

Their biggest clients are hospitals and insurers. There are people called "frequent flyers" who come to the emergency room frequently and while they might only be 2% of clients they account for a huge share of costs. It's actually cheaper to hire social workers to provide these people with the things they actually need (i.e. treating the causes) instead of the symptoms.

Here's an example. Guy is low income. Cannot afford food & insulin for diabetes. So he chooses to not buy insulin and the food he does buy is high in sugar (think cheap whitebread). Goes into diabetic shock and gets sent to the emergency room. Instead it's much cheaper for everyone involved if a social worker finds out the problem, helps him research some foodbanks to cover his nutritional needs or gets him on a discount insulin program. It's not free, but it's much cheaper than trying to clean up a mess after the fact.

Post reply on HN