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

This is silly. The biggest social programs in America are Social Security (basically a UBI for old folks), Medicare (health care), and the #1 economic development success of all time, public education. The kind of social programs they're talking about in this article are so unambitious it's hardly surprising their results aren't measurable.

You are 100% correct if you have zero context. Let me provide some context.

If you look at https://80000hours.org/about/ their goal is to help people make the world better, by a combination of career and charitable donations.

> Our aim is to help as many people as possible lead high-impact careers.

> We do this by providing career advice for talented young people who want to have a social impact.

Given this context, several questions arise:

1. What to give to?

2. Which areas to give to? Health? Social? Education?

3. Which countries?

4. Which programs that you can give to have positive outcome? How do you decide which to give to?

That's the context of question being asked.

The specific article is about an issue that, in this community, is a big question, and 80,000 hours (a YC non-profit BTW) tries to quantify the truth of this common - in their sphere - belief.

It is not an attack, or a denial of any of the things you stated, just a different context for inquiry.

This is a large problem in Canada as well. A relative is an ER nurse and she has endless amounts to say about the number of people who come in for these kinds of reasons, or less. The hospital she works at is in a town that was economically decimated in the late 80's and only just started to recover. The number of people who go into emergency because of alcoholism, self-neglect, loneliness, and of course (probably encompassing the rest) mental illness is staggering. She says it's one thing to know their names, but it's bad when you know how their [grand-/]children are doing in school, and who visited or did not visit last week. She is not shy about how much she thinks that social programs need to be expanded in accessibility. I agree.

From what I understand, much of the medical world feels the same way. That's one thing I would certainly add to Canadian health care -- some kind of minimum of public mental health and social work. There are a lot of programs already that don't attach a fee for the patient/subject, but they're not necessarily common knowledge or accessible to people as a "walk in and book" sort of service.

edit: To wit, I think the reasons social programs don't work is inaccessibility to those who really need them, lack of familiarity with them, and taboo (but this last one is probably another discussion). I think right now we need more of them, and they need to have access to proper resources and that their end benefit to society would be substantial. I'd be interested in seeing some projections surrounding that to be sure.

People are failing to recognize that it doesnt matter what percentage of programs fail/succeed. The metric that matters is the ROI on programs. If 25% of programs succeed, giving a 100x ROI, then they more than make up for the failed programs. Unless we know a better way (UBI seems promising) to help people, then programs seem to be the best bet.

The problem is that people hear "75% of programs fail" and think "Dont start programs". This is like saying "75% of businesses fail therefore do not start a business" .

These also happen to be the most difficult to use in terms of dosing frequency and rate of blood glucose control.

In the US of all places, there isn't a good argument for the very, very slightly modified newer insulins to cost so much more.

(these points are alluded to in the article, just want to mention them here, as well)