Epsilon walks up to you with two boxes, A and b, labeled in rather childish-looking handwriting written in crayon.

"In box A," he intones, sounding like he's trying to be foreboding, which might work better when he hits puberty, "I may or may not have placed a million of your human dollars."  He pauses for a moment, then nods.  "Yes.  I may or may not have placed a million dollars in this box.  If I expect you to open Box B, the million dollars won't be there.  Box B will contain, regardless of what you do, one thousand dollars.  You may choose to take one box, or both; I will leave with any boxes you do not take."

You've been anticipating this.  He's appeared to around twelve thousand people so far.  Out of eight thousand people who accepted both boxes, eighty found the million dollars missing, and walked away with $1,000; the other seven thousand nine hundred and twenty people walked away with $1,001,000 dollars.  Out of the four thousand people who opened only box A, only four found it empty.

The agreement is unanimous: Epsilon is really quite bad at this.  So, do you one-box, or two-box?

There are some important differences here with the original problem.  First, Epsilon won't let you open either box until you've decided whether to open one or both, and will leave with the other box.  Second, while Epsilon's false positive rate on identifying two-boxers is quite impressive, making mistakes about one-boxers only .1% of the time, his false negative rate is quite unimpressive - he catches 1% of everybody who engages in it.  Whatever heuristic he's using, clearly, he prefers to let two-boxers slide than to accidentally punish one-boxers.

I'm curious to know whether anybody would two-box in this scenario and why, and particularly curious in the reasoning of anybody whose answer is different between the original Newcomb problem and this one.

Over the last few days I've been reviewing the evidence for EA charity recommendations. Based on my personal experience alone, the community seems to be comprehensively inept, poor at marketing, extremely insular, methodologically unsophisticated but meticulous, transparent and well-intentioned. I currently hold the belief that EA movement building does more harm than good and that is requires significant rebranding and shifts in its informal leadership or to die out before it damages the reputation of the rationalist community and our capacity to cooperate with communities that share mutual interests.

It's one thing to be ineffective and know it. It's another thing to be ineffective and not know it. It's yet another thing to be ineffective, not know it, yet champion effectiveness and make a claim to moral superiority.

In case you missed the memo deworming is controversial, GiveWell doesn't engage with the meat of the debate, and my investigations of the EA community's spaces suggests that it's not at all known. I've even briefly posted about it elsewhere on LessWrong to see if there was unspoken knowledge about it, but it seems not. Given that it's the hot topic in mainstream development studies and related academic communities, I'm aghast at how irresponsive 'we' are.

What's actionable for us here. If you're looking for a high reliability effective altruism prospect, do not donate to SCI or Evidence Action. And by extension, do not donate to EA organisations to donate to these groups, including GiveWell. I am assuming you will use those funds more wisely instead, say buying healthier food for yourself.

For who don't to review the links for a more comprehensive analyses from Cochrane and GiveWell, here is one summary of the debate recommended in the Cochrane article:

Last month there was another battle in an ongoing dispute between economists and epidemiologists over the merits of mass deworming. In brief, economists claim there is clear evidence that cheap deworming interventions have large effects on welfare via increased education and ultimately job opportunities. It’s a best buy development intervention. Epidemiologists claim that although worms are widespread and can cause illnesses sometimes, the evidence of important links to health is weak and knock-on effects of deworming to education seem implausible. As stated by Garner “the belief that deworming will impact substantially on economic development seems delusional when you look at the results of reliable controlled trials.”

Aside: Framing this debate as one between economists and epidemiologists captures some of the dynamic of what has unfortunately been called the “worm wars” but it is a caricature. The dispute is not just between economists and epidemiologists. For an earlier round of this see this discussion here, involving health scientists on both sides. Note also that the WHO advocates deworming campaigns.

So. Deworming: good for educational outcomes or not?

On their side, epidemiologists point to 45 studies that are jointly analyzed in Cochrane reports. Among these they see few high quality studies on school attendance in particular, with a recent report concluding that they “do not know if there is an effect on school attendance (very low quality evidence).” Indeed they also see surprisingly few health benefits. One randomized control trial included one million Indian students and found little evidence of impact on health outcomes. Much bigger than all other trials combined; such results raise questions for them about the possibility of strong downstream effects. Economists question the relevance of this result and other studies in the Cochrane review.

On their side, the chief weapon in the economists’ arsenal has for some time been a paper from 2004 on a study of deworming in West Kenya by Ted Miguel and Michael Kremer, two leading development economists that have had an enormous impact on the quality of research in their field. In this paper, Miguel and Kremer (henceforth MK) claimed to show strong effects of deworming on school attendance not just for kids in treated schools but also for the kids in untreated schools nearby. More recently a set of new papers focusing on longer term impacts, some building on this study, have been added to this arsenal. In addition, on their side, economists have a few things that do not depend on the evidence at all: determination, sway, and the moral high ground. After all, who could be against deworming kids?

 

---

 

Additional criticisms of GiveWelL charities: http://lesswrong.com/lw/mo0/open_thread_aug_24_aug_30/cp8h

The kind of work I think EA's should be focussing on http://lesswrong.com/lw/mld/genosets/cnys AND

http://lesswrong.com/r/discussion/lw/mk2/lets_pour_some_chlorine_into_the_mosquito_gene/

The problem with MIRI: http://lesswrong.com/lw/cr7/proposal_for_open_problems_in_friendly_ai/cm2j