The most difficult part of this worldview for me to reconcile has been to convince myself that the subset of negative-impact health interventions can have a large enough magnitude of an effect to counteract the health interventions that have a strong evidence base of a positive effect.
…
do you think that some educational interventions have a negative impact, and thus offset positive effects elsewhere? If so, what is the nature of those negative-effect interventions?
1. The commenter does not cite Robin Hanson, so I must make sure that readers are familiar with this paper.
2. I think that in both health care and education, other factors affect outcomes a great deal.
3. In the case of health care, individual genes and behavior, along with public health and cultural trends, are very important. Health care providers work on very small margins. Sometimes they make things better, and sometimes they make things worse. To justify the huge resources that we spend on health care, I think you have to value the occasional benefits very highly and assume that those resources could not be used more effectively on individual and collective efforts aimed at prevention.
4. In the case of education, individual genes and behavior, along with the overall cultural environment, are very important. Educators work on very small margins. Sometimes they make things better, and sometimes they make things worse.
5. Do the interventions that make things work exactly offset those that make things better? I think not. I am optimistic that health care providers and educators do more good than harm. But I think that interventions work on small margins, so that the average benefits turn out to be insignificant relative to the costs.
Harmful interventions:
Ending tracking
De-emphasizing demonstrated test scores on difficult tests in favor of grades.
Increased legal protections for discipline disasters.
Of the first two, neither have done anything to improve achievement or access. Both have done tremendous damage to high achievers. Caveat: most people think of helping high achievers via coverage (learning faster) but in fact, this has been the most damaging aspect of the changes. Kids are only able to demonstrate ability by the age at which they take a particular math course, and so you have sophomores taking 1st year calculus classes and memorizing enough to pass an AP test without particularly understanding the underlying math.
No, the cost to high achievers is that we aren’t pushing them hard because we can’t, because GPA means so much that teachers can’t create a proper leveling schema. I taught trig. Five kids never showed up regularly, so they got Fs. Other kids worked hard despite having no understanding of the concept in a class they had no choice but to take, so they got Ds. Others had some limited understanding and could successfully do about half the work, so there’s your C. And so on.
And that’s in math. Things are far worse in literature and history, so that even top students are often terrible writers and teachers can’t even begin to address it in classes where the bottom students can’t even read and everyone’s using Schmoop or Sparcnotes to get the short version. We can’t have classes just for bright kids and give hard workers of merely adequate performance in a tough class a C for trying, because that C will do tremendous damage to those kids without the context of a test score.
The third is causing tremendous damage in low income schools, as well as creating more segregation as parents who can leave do. (I get annoyed at people who blame teachers for reduced discipline. It’s a specific policy demand forced on us by the state.)
Costly interventions:
Special education now gives additional money to 1 out of 8 kids and we see nothing for it. Special ed means different things. (1) The severely retarded, who cost hundreds of thousands to educate and should not be part of public school. It’s free childcare service for 16 years, at which point we then dump the kid over to a different state budget. (2)Then there’s the low IQ kids, what we always meant by special ed, who get relatively little in services. (3) Then there’s the emotionally damaged kids who can’t function in regular classes despite no IQ problems. (4) Then there’s the kids with a real disability (blind, wheelchair bound) who get access and aids. (5) Then there’s everyone with a “learning disability”–the fastest growing group. Only 2 and 4 were originally intended by the special ed category. We should dump 5 entirely, create centralized institutions for 1 and 3. 2 and 4 would still cost a lot, but at least they were intended to be addressed by K-12 ed.
We spend billions on “English language instruction”, which has’t one meaning. In schools like mine, it means free English lessons for immigrant kids who just got here–mandatory lessons that are often frustrating to bright kids whose English is adequate to work in academic courses (and far better than the bottom third in each course) but aren’t allowed because the “get me out of ELL” score is ridiculously high. In more homogenous schools, it means running half the classes in (usually) Spanish, because ELL from the 60s on was designed on the expectation that ELL kids would be illegal immigrants from across the border.
So the things that do actual harm aren’t extremely expensive, and the things that are a waste of time don’t so much do harm as create hugely expensive systems to support kids that aren’t improving their education and often given tremendous support to kids who just got here while not offering that service (enrichment) to citizens who might benefit from it.
I wrote about some of these in the links in this piece here: https://educationrealist.wordpress.com/2015/07/31/five-education-policy-proposals-for-2016-presidential-politics/
Clearly GPA should be replaced by class rank %. Then SAT. And finally AP and aptitude tests. I’m almost dubious that they wouldnt be doing such no-brainer things…which makes me almost sure they aren’t. So why aren’t they doing thsee type things?
A common core curriculum but not standardized aptitude is the exact wrong way to go.
Oh, and they do balls for career aptitude. That is what I call a tell.
I got to college and said….hmmm….I don’t know, I guess I’ll take engineering. Good choice, mostly lucky.
The solutions are obvious and not even all that complicated. The problem is changing the institutional dysfunction. That is the only real problem.
A great summary comment. I wish some “medicine realist” doctor could make as efficient a litany.
I’m going to propose we grade doctors by periodically testing the health of their patients.
They all die.
I suspect your comment was meant (somewhat) glibly, since if you measured on the health of the patient alone then doctors would simply try to avoid dealing with people who are in chronically poor health and will never be “very healthy”. Measuring the change in health would be better.
Furthermore, if you measured “change in health” of the patient against amount of treatment provided, I think you would have a pretty good objective measure of a doctor’s effectiveness.
If we knew how to measure health or education we’d know how to produce it.
I unfortunately don’t have a “medicine realist” to point you to, but after having talked with medical professionals and people in the insurance industry, here’s what I’ve gathered as important:
Insurance:
1. Encourage HSAs and high-deductible health plans, perhaps with government contributions for those with low incomes. Most people don’t have “health insurance”; they have concierge medicine, which is going to be more expensive. This also incentives people to take care of their health and encourages price transparency and competition for many basic medical services.
2. Allow insurance companies to compete across state lines.
Medical Training:
3. Have many tiers of medical professionals, so that you are not going to a very expensive doctor for every ache and pain when you could go to a less expensive nurse practitioner or chiropractor. (We actually have this to some extent, but we should encourage it further)
4. Have the tiers build on each other, so that even if you aren’t able to complete a full M.D. program (for instance) you can still find meaningful employment.
5. Remove the requirement for an undergraduate degree before attending medical school. Follow Ireland’s model of a 5 year medical school where the first year is all the pre-requisites you would normally study in college.
R&D:
6. Change the intellectual property scheme for R&D development so that it is less restrictive but lasts longer. Perhaps through a system of granting manufacturing licenses rather than patents.
7. For FDA approval, only require proof of safety, not effectiveness for new medical treatments.
Two unrelated comments:
(1) Medical licensing is a tricky balance. There are obvious benefits to ensuring a baseline level of competence. Especially since outcomes are so poorly measured. But the current limits on provider supply drive up cost. Not least because the provider sector is structured around the needs of the physician, not the patient or payer. Which drives up cost and limits innovation.
(2) Healthcare is very much another industry that suffers from educationrealist’s list of costly interventions. 5% of patients drive 50% of cost, and some of that is from populations that are structurally higher cost. There’s a slippery slope, along which we find drivers that the individual has no ability to control (think expensive congenital defects), drivers that can be influenced by individual behavior (diabetes or CHF), and technology that has been developed and that as a society we are unwilling to withhold (expensive end of life care, cancer drugs with marginal impact).
Conceptually, I like high-deductible HSAs as a way to distinguish between low- and high-value care. But the rub comes in separating the 80% of the population for whom they would work from the 20% (or 5%) for which a subsidy would be appropriate.
Nassim Taleb discusses iatrogenics in his Anti-Fragile. I’m going to have to go back and read that section again but with education in mind. It’s only been in the last century that going to the doctor, i.e., after penicillin, was a net positive on balance, before going to the doctor/hospital increased your likelihood of death. Death from hospital-sourced infection is still very high.
In regards to education, there is ample evidence that school destroys creativity in a child by 3rd grade. That many schools induce “school helplessness”, i.e., lack of initiative with school tasks. Some students thrive in a strict “sage on the stage” method, while others bloom in self-directed programs. And there is very little energy applied by the educators to pair the student with their best method. It seems that for education the iatrogenics has flipped over the last century to a net negative, beyond grammar school.
“In the case of tonsillectomies, the harm to the children undergoing unnecessary treatment is coupled with the trumpeted gain for some others. The name for such net loss, the (usually hidden or delayed) damage from treatment in excess of the benefits, is iatrogenics, literally, “caused by the healer,” iatros being a healer in Greek. We will posit in Chapter 21 that every time you visit a doctor and get a treatment, you incur risks of such medical harm, which should be analyzed the way we analyze other trade-offs: probabilistic benefits minus probabilistic costs.”
Taleb, Nassim Nicholas (2012-11-27). Antifragile: Things That Gain from Disorder
Before comments close, Taleb’s _Black Swan_ has a nice couple of pages about expertise. He essentially says “some fields don’t have experts.”
soil scientists and civil engineers who design bridges are experts.
He didn’t think much of guidance counselors and career counselors, nor of psychologists providing therapy. He said that they were in fields that essentially did not have experts.
Some find his tone offensive, but I often find it adds to his already valuable insights.
The notion of expertise is still dimly understood.
Someone else said that an emergency room nurse or a fire watch captain with 20 years experience has some expertise, in the sense that if they suddenly get worried (about the patient or the fire), you should listen.
The underlying issue might be “domain.”