Conflict-poverty

	B	C	D	E	F	G
1
2
3	Model for linking poverty/GDP with probability of conflict
4	This is the Collier model (which perhaps might be the same as the Collier-Hoeffler model)
5	The formulation used here is as per the quantification provided in Collier's book The Bottom Billion
6	It appears to be linked to the models set out in this World Bank paper by Collier and Hoeffler, which modelled war in Africa http://documents.worldbank.org/curated/en/822561468768022483/pdf/28128.pdf
7	More about the rebel-greed hypothesis is set out here: https://www.jstor.org/stable/174593?seq=1
8	This paper also appeared relevant: https://www.econ.nyu.edu/user/debraj/Courses/Readings/CollierHoeffler.pdf
9		Collier	Risk of civil war in a typical low income country in 5 years	14%	The Bottom Billion, Chapter 2, page 20
10		Collier	Improvement in war probability for each %age point of growth	1%	The Bottom Billion, Chapter 2, page 20
11
12	Model excluding tail risk (based on a putative poverty reduction intervention in Kenya)
13
14		Impact of Giving on GDP	Population of Kenya	49,700,000	Source: whatever comes up first when googling "Population of Kenya"
15		Impact of Giving on GDP	Proportion that is poor	36.80%	Source: world bank data https://www.indexmundi.com/facts/indicators/SI.POV.DDAY/compare?country=ke
16		Impact of Giving on GDP	Population on a low income in Kenya	18,289,600
17		Impact of Giving on GDP	Cost per person	$1,000	GiveDirectly intervention involves giving c$1000 to each poor person.
18		Impact of Giving on GDP	Total cost	$18,289,600,000	Unmodelled: admin costs -- if this happened at scale, maybe would need to invest more in admin, e.g. maybe more effort would be needed to verify identity
19		Impact of Giving on GDP	Rate of return	10%	Source: GiveWell cost effectiveness analysis, 'User Inputs' tab; https://www.givewell.org/how-we-work/our-criteria/cost-effectiveness/cost-effectiveness-models
20		Impact of Giving on GDP	Improvement in GDP per capita from investment return	$100
21		Impact of Giving on GDP	Assumed time period over which $1000 of capital is spent	10	years
22		Impact of Giving on GDP	Improvement in GDP per capita from spending cash	$100
23		Impact of Giving on GDP	Add in factor for impact on currency?		unmodelled
24		Impact of Giving on GDP	Improvement in GDP per capita	$200
25		Impact of Giving on GDP
26		Impact of Giving on GDP	Kenya-wide GDP per capita	$1,711	https://data.worldbank.org/indicator/NY.GDP.PCAP.CD?locations=KE
27		Impact of Giving on GDP	Improvement in gdp per cap	$73.60
28		Impact of Giving on GDP	% improvement in gdp	4.30%	per annum
29
30		Reduction in prob(war)	Reduction in probability of war over next 5 years	4.30%
31		Reduction in prob(war)	Discount rate	5%
32		Reduction in prob(war)	No of years between wars	5	The Bottom Billion, Chapter 2, page 20
33		Reduction in prob(war)	Modified discount rate	27.63%
34		Reduction in prob(war)	Present value	0.16	wars averted
35
36
37		No of deaths	No of battle-related deaths over 1989-2018	1,408,882	Source: took the data from Uppsala (UCDP) and summed column M https://ucdp.uu.se/downloads/index.html#battlerelated
38		No of conflicts	No of conflicts over 1989-2018	179	Source: took the data from Uppsala (UCDP) and counted distinct conflicts from column A https://ucdp.uu.se/downloads/index.html#battlerelated
39		Deaths per conflict	Deaths per conflict 1989-2018	7,871
40
41	Excl tail risk	Cost per death averted	Expected no of battle deaths averted because of poverty intervention	1,226	Ignores tail nuclear risk
42	Excl tail risk	Cost per death averted	Total cost of intervention	$18,289,600,000	Ignores tail nuclear risk
43	Excl tail risk	Cost per death averted	Cost per death averted	$14,920,178	Ignores tail nuclear risk
44
45	Model including tail risk (based on a putative poverty reduction intervention in India and Pakistan)
46
47	Tail risk	India	Population on a low income in India	50,000,000	https://timesofindia.indiatimes.com/india/new-data-may-show-big-cut-in-number-of-poor/articleshow/67705787.cms
48	Tail risk	Pakistan	Poverty headcount ratio at $1.90 a day (2011 PPP) (% of population)	3.90%	http://povertydata.worldbank.org/poverty/country/PAK
49	Tail risk	Pakistan	Population, total	212,200,000	http://povertydata.worldbank.org/poverty/country/PAK
50	Tail risk	Pakistan	Population on a low income in Pakistan	8,275,800
51	Tail risk	India and Pakistan	Population on a low income	58,275,800
52	Tail risk		Cost to raise all out of poverty	$58,275,800,000
53	Tail risk	Expected deaths in nuclear scenario	Prob(at least 1 million dead)	30%	Source: https://www.fhi.ox.ac.uk/reports/2008-1.pdf (also mentioned here https://forum.effectivealtruism.org/posts/PAYa6on5gJKwAywrF/how-likely-is-a-nuclear-exchange-between-the-us-and-russia)
54	Tail risk	Expected deaths in nuclear scenario	Prob(at least 1 billion dead)	10%	Source: https://www.fhi.ox.ac.uk/reports/2008-1.pdf (also mentioned here https://forum.effectivealtruism.org/posts/PAYa6on5gJKwAywrF/how-likely-is-a-nuclear-exchange-between-the-us-and-russia)
55	Tail risk	Expected deaths in nuclear scenario	Prob(human extinction)	1%	Source: https://www.fhi.ox.ac.uk/reports/2008-1.pdf (also mentioned here https://forum.effectivealtruism.org/posts/PAYa6on5gJKwAywrF/how-likely-is-a-nuclear-exchange-between-the-us-and-russia)
56	Tail risk	Expected deaths in nuclear scenario
57	Tail risk	Expected deaths in nuclear scenario		Probability	Expected number of deaths (geometric average)
58	Tail risk	Expected deaths in nuclear scenario	1 million to 1 billion dead	20%	31,622,777
59	Tail risk	Expected deaths in nuclear scenario	1 billion to 7 billion dead	10%	2,645,751,311
60	Tail risk	Expected deaths in nuclear scenario	7 billion dead	1%	7,000,000,000	Note: this gives no credit to people not born yet
61	Tail risk	Expected deaths in nuclear scenario	Expected number of deaths (calc)	340,899,686	Note: this assumes that the deaths predicted in the survey relate to one conflict. On balance this seems reasonable, given that the other surveys mentioned in the Luisa Rodriguez post (https://forum.effectivealtruism.org/posts/PAYa6on5gJKwAywrF/how-likely-is-a-nuclear-exchange-between-the-us-and-russia) seem to give a probability of incidence which is consistent with that of the FHI 2008 survey used here
62	Tail risk	Expected deaths in nuclear scenario	Adjusted for the fact our model is not global	34089968.64	The calculation above is based on estimates of the global impact of nuclear war, and this adjustment is to account for the fact that the poverty intervention is imagined to happen in one country
63	Tail risk
64	Tail risk	Cost per death averted	Expected no of battle deaths averted because of poverty intervention	5,309,173	Ignores tail nuclear risk
65	Tail risk	Cost per death averted	Total cost of intervention	$58,275,800,000	Ignores tail nuclear risk
66	Tail risk	Cost per death averted	Cost per death averted	$10,976	Ignores tail nuclear risk
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