I'm a doctor working towards the dream that every human will have access to high quality healthcare. I'm a medic and director of OneDay Health, which has launched 53 simple but comprehensive nurse-led health centers in remote rural Ugandan Villages. A huge thanks to the EA Cambridge student community in 2018 for helping me realise that I could do more good by focusing on providing healthcare in remote places.
Thanks for the update, and the reasons for the name change make s lot of sense
Instinctively i don't love the new name. The word "coefficient" sounds mathsy/nerdy/complicated, while most people don't know what the word coefficient actually means. The reasoning behind the name does resonate through and i can understand the appeal.
But my instincts are probably wrong though if you've been working with an agency and the team likes it too.
All the best for the future Coefficient Giving!
Thanks @mal_graham🔸 this is super helpful and makes more sense now. I think it would make your argument far more complete if you put something like your third and fourth paragraphs here in your main article.
And no I'm personally not worried about interventions being ecologically inert.
As a side note its interesting that you aren't putting much effort into making interventions happen yet - my loose advice would be to get started trying some things. I get that you're trying to build a field, but to have real-world proof of this tractability it might be better to try something sooner rather than later? Otherwise it will remain theory. I'm not too fussed about arguing whether an intervention will be difficult or not - in general I think we are likely to underestimate how difficult an intervention might be.
Show me a couple of relatively easy wins (even small-ish ones) an I'll be right on board :).
Great answert thanks. Makes sense on the selection. I agree its good to leave out technical assistance orgs on a first pass - super hard to know what to attribute.
On the survey front I'm not so concerned about the bias of enumerators (although this can be a factor), its about inflation of effects from people in the program, and sometimes even deflation from those not. These programs are such a big deal, that everyone knows that the program has been going on, and what the phone calls will be about, even if the programs aren't mentioned. People are smart and observant - sometimes we underrate just how good communication networks are. Someone in the village will know the people in town running the program and a distant relative will be involved etc. That goes for the treatment and control villages. Its the rational and correct play to pump up any program that comes to your village, because you correctly think that both you and surrounding communities are more likely to get more help in future future. I call this "future hope" bias in surveys.
The problem with mixing survey and objective measures is that the survey can then bias your recording of the other "objective measures" and vice versa. Even the in-person verification of surveys can be gamed to some extent by showing people your neighbours chickens or gardens. Have seen that before :D. All things equal of course it would be great to have both surveys and objective measures, but as soon as you are asking lots of questions, you're going to be subconsciously/consciously backing those original surveys up with your "objective findings". Surveys are also much slipperier to check than other methods. There are ways to test them statistically but these are limited, its not like Evidence Action's chlorine dispensers where you can just go back and check whether the water has chlorine.
I think that consumption does have decent, if not perfect objective measures which can be quickly checked physically. Checking this stuff can take similar amounts of time to in-person discussions and surveys sometimes.
- School term reports + school fee receipts
- Amount of food/charcoal/salt/sugar/tea in the house and other consumables
- Evidence of recent construction or home renovation.
- Medical bills and receipts
- Airtime and mobile money use history recorded on phones
- Weight by height of kids (stunting measures)
Livestock and Gardens are OK but can be gamed far easier than stuff which is inside their actual home itself
Earlier on in cash transfer days they used physical health objective measures more, but now they use them less for non-invasive reasons.
Even these seemingly "objective" measures will be somewhat biased by the enthusiasm and attitude of the household members showing researchers these things and putting them forward. For example someone in the program is more likely to bring all their food out to see, whereas someone who isn't might be more ambivalent.
Keep in mind I'm not criticising the programs at all here - I'm a huge fan. I think the survey issue is an even bigger problem for cash transfers. GiveDirectly surveys are far more likely to be off-base than surveys these programs, because people will feel far more straightforwardly positive about cash gifts than this kind of program. I think surveys are super useful and often the only practical option, but I'm far more skeptical about effect sizes of surveys than the average researcher, having seen a few surveys in action here in Uganda, and even participated in in-house surveys for my org OneDay Health. Its far more tricky and fraught than it can seem and I'm fairly confident in general surveys in general regularly inflate real effects by some X%.
I completely agree. Is there any argument that an IPO good be good for AI safety? Would be interesting to hear it.
Love this post. Graduation programs really be banging right now!
It seems the main reason why you picked them over other poverty graduation orgs like village enterprise is the low cost rollout which leads to better cost-effectiveness? Was wondering if you might elaborate a bit more on precisely what attracted you to them over say village enterprise or BRAC or others.
Also as a slightly unnecessary sidenote, Judging RCT results just by surveys is a bugbear of mine, and this RCT like basically all others in the livelihoods space is largely based on surveys. I would kind of love to see an RCT on cash transfers or graduation programs which JUST used objective methods (household goods, school reports, food at home, cash at hand etc.) and zero survey because I still think the effect of these interventions might be quite inflated by survey effect. But we probably won't convince someone to do a study and just get way less info which would be the case without surveys.
I think this is a factor for sure, but likely to be dwarfed by the stock options / kidding yourself factors that @Ian Turner and @MichaelDickens have put forward. Also I think there's a trajectory where you join for mission/leadership, then kind of realise that what you are building could kill everyone then self justify with something along the lines of "we are the good guys" with the devil of financial incentives sitting on one shoulder as well.
At least that's has somewhat paralleled my experience of periods of eating factory farmed meat and other moral failures in my life.
Of these 5, I would expect 1 and 2 to be "fake objections" much of the time, wheras 3 to 5 can often be meaninfgul concerns. Just because you can't resolve a question doesn't mean its necessarily a bad question or one that should be ignored. I agree if you think deeply and decide that you're just making excuses for yourself and living in fear, then these aren't real questions on a personal level. But they are real for many.
'm in a critical role in global health.[5] If I leave, the work won't get done and people will die. Plus, too many people are already leaving the field.
I agree with you in humans, but who knows how these degrees of pain might work in animals? I think with these degrees of uncertainty in animals it's not unreasonable to have smaller multipliers
Not my area, but a well written concise post, with great explanation. How right you are I do not know (Not my area again) but I appreciate clear arguments that can easily be engaged with.
As giving season is here, lets do our best to not fall into this trap. I don't agree this problem is the norm, but I've seen it both in myself and others.
"Perhaps the most overrated virtue in our list of shoddy virtues is that of giving. Giving builds up the ego of the giver, makes him superior and higher and larger than the receiver. Nearly always, giving is a selfish pleasure, and in many cases it is a downright destructive and evil thing.
One has only to remember some of our wolfish financiers who spend two-thirds of their lives clawing fortunes out of the guts of society and the latter third pushing it back. It is not enough to suppose that their philanthropy is a kind of frightened restitution, or that their natures change when they have enough. Such a nature never has enough and natures do not change that readily. I think that the impulse is the same in both cases.
For giving can bring the same sense of superiority as getting does, and philanthropy may be another kind of spiritual avarice."
Steinbeck -The Log from the Sea of Cortez