It seems alarming that GiveWell bases their significant donation recommendations only on one study[1] that, furthermore, does not seem to understand beneficiaries’ perspectives but rather estimates metrics that relate to performance within hierarchies that historically privileged people set up: school attendance[2], hours worked[3], and income.
GiveWell’s reports should align more closely with academic norms where authors are expected to fully explain their data, methods, and analysis, as well as the factors that their conclusions are sensitive to
I disagree that GiveWell’s reports should align more closely with academic norms, because these norms do not engage intended beneficiaries.
Explanations can help differentiate the actually most helpful programs from those made prestige by big/small numbers and convoluted analyses.
Allowing GiveWell’s audience tweak the factors and see how conclusions change would show the organization’s confidence in their (moral) judgments.
‘Data’ should not be confused with ‘numbers.’ Focus group data may be invaluable compared to quantitative estimates when a solution to a complex problem is being found.
The only evidence GiveWell uses to estimate the long-term effects of deworming comes from a study of the Primary School Deworming Project (PSDP) using the Kenya Life Panel Survey (KLPS) (Miguel & Kremer, 2004) and its follow-ups (Baird et al., 2016; Hamory et al., 2021). (HLI, Appendix: Calculations of Deworming Decay )
I am unsure whether unpaid domestic and care work was considered within hours worked—excluding this would imply greater value of paid over unpaid work, a standard set up by the historically privileged.
It seems alarming that GiveWell bases their significant donation recommendations only on one study[1] that, furthermore, does not seem to understand beneficiaries’ perspectives but rather estimates metrics that relate to performance within hierarchies that historically privileged people set up: school attendance[2], hours worked[3], and income.
I disagree that GiveWell’s reports should align more closely with academic norms, because these norms do not engage intended beneficiaries.
Explanations can help differentiate the actually most helpful programs from those made prestige by big/small numbers and convoluted analyses.
Allowing GiveWell’s audience tweak the factors and see how conclusions change would show the organization’s confidence in their (moral) judgments.
‘Data’ should not be confused with ‘numbers.’ Focus group data may be invaluable compared to quantitative estimates when a solution to a complex problem is being found.
School curricula in developing contexts may include post-colonial legacy, select elites while leaving most behind, or optimize for raising industrial workforce that may prevent global value chain advancement of industrializing nations but make the countries an instrument for affordable consumption of foreign-made goods.
I am unsure whether unpaid domestic and care work was considered within hours worked—excluding this would imply greater value of paid over unpaid work, a standard set up by the historically privileged.