Imagine Worldwide: Public

	A	B	C	D
1	Variable	Value	Source	Notes
2
3	Costs
4	Total Philanthropic Cost	$174,670,784	https://docs.google.com/spreadsheets/d/1f0UxVPYJdEfI1sVIcMD1btWEJalzzxdOB0eDa8C-HVU/edit#gid=134296518	Imagine Worldwide projections, sanity checked by looking at the cost to provide a tablet amortized over 5 years and annual maintenance costs for one child. Conservatively assumes that spending on implementation and hardware does not benefit the government phase of the program, and that implementation costs generally persist on an annual basis.
5	Total number of student-years during ramp-up	11,717,043	https://docs.google.com/spreadsheets/d/1f0UxVPYJdEfI1sVIcMD1btWEJalzzxdOB0eDa8C-HVU/edit#gid=134296518	Summing the number of students served in each of the first six years of implementation. This is during the philanthropic funding phase, before the government takes over.
6	Philanthropic cost per student-year	$14.91	calculation
7
8	Government Phase
9	Additional expected years of effective operation after government takeover	10	subjective, based on assumption that effectiveness of program will decay to that of other government spending on education over 20 years. We do not do any time discounting because both costs and benefits are in the future.
10	Total students served per year	3,446,189	https://docs.google.com/spreadsheets/d/1f0UxVPYJdEfI1sVIcMD1btWEJalzzxdOB0eDa8C-HVU/edit#gid=134296518
11	Probability of government takeover happening	0.5	subjective	Based on the assumption that there is a 10%-20% chance that each of the following may lead Malawi to cancel their plan to take over implementation nationwide: 1) IW is not able to get enough funding, or otherwise fails to implement the program in all schools, 2) IW is not able to get sufficient learning gains as it scales up, 3) The leadership or funding situation in Malawi changes sufficiently for them to renege on their committment as the UN General Assembly, 4) Something unforeseen happens.
12	Government cost per student-year after takeover	$10.44	https://docs.google.com/spreadsheets/d/1f0UxVPYJdEfI1sVIcMD1btWEJalzzxdOB0eDa8C-HVU/edit#gid=134296518
13	Total Government Cost	$179,873,980	calculation
14
15	Impacts
16	Initial standard deviation impact on learning gains per year	0.33	https://docs.google.com/spreadsheets/d/1f0UxVPYJdEfI1sVIcMD1btWEJalzzxdOB0eDa8C-HVU/edit#gid=1879027745	Weighted Average effect size from 6 studies tracking impact on math and literacy scores, weighted by relevance to current program.
17	Learning gains adjusted for sample size and reduced effectiveness at scale-up	0.22	calculation	Even though our general recommendation would be only a 25% weight on the prior of 0.1 given the adjusted sample size >2000, we increase the weight on the prior to 50% because this is a national scale-up with some government involvement. The adjustment is based on the recommendation in Section III.A of the Education recommendations report.
18	Units of ln(income) per 1 SD increase	0.989	https://docs.google.com/spreadsheets/d/1V-Dk9VeaRaSgbak9C-Ube1paUALtPASByZ_DdiMUX78/edit#gid=1236602313
19	Value Assigned to 1 unit of ln(income)	2.74	Moral Weights tab
20	WELLBYs Per Student	0.58	calculation
21	Total Value of Learning Impact generated during ramp-up	6,821,232	calculation
22	Total Value of Learning Impact generated after government takeover	10,031,223	calculation
23	Total Value of Learning Impact	16,852,455	calculation
24
25	Wellby per government education dollar spent	0.004951	https://docs.google.com/spreadsheets/d/1Ut1EurGgvjdPf1pdW7TCU5FJ4SzIbJ_rvplWQJ_6vPw/edit#gid=1158468609	Counterfactual non-philanthropic tab
26	Units of value that could have been generated counterfactually by government spending	890,556	calculation
27
28	Total value generated net of counterfactual government value	15,961,899	calculation
29
30	Cost Effectiveness
31	Value generated per philanthropic dollar before accounting for government phase	0.0391	calculation
32	Value generated by GiveDirectly per philanthropic dollar	0.00586	Results Tab of internal Current Gen Sheet
33	Multiple of Givedirectly before leverage	6.7	calculation
34
35	Value generated per philanthropic dollar after accounting for government funding	0.0914	calculation
36	Value generated by GiveDirectly per philanthropic dollar	0.00586	Results Tab of internal Current Gen Sheet
37	Multiple of Givedirectly after leverage from government spending	15.6	calculation
38
39	Discount for counterfactuality of funding	0.7	subjective	Based on IW's success raising funds for the first year of their program, and external anecdotal information about their connections in the philanthropic sector. We assume that there is some chance that money ends up counterfactually increasing the size of less effective IW programs because the Malawi program would have been fully funded without FP.
40	Multiple of Givedirectly after discounting for counterfactuality	10.9	calculation
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100