ABCDEFGHIJKLMNOPQRSTUVWXYZ
1
2
Arbitrary cohort (age 15+)10,000
3
4
Alcohol units of consumption per person per year228
1 litre is 40 units. 2 units = 1 beer. WHO estimates 5.7 litres of consumption per capita in India (India chosen arbitrarily). I expect this overestimates the relationship between alcohol and burden of disease because current burden of disease is partially based on past alcohol consumption and alcohol consumptiuon is rising in India. I haven't looked at past consumption patterns but guess 4 litres per capita for the base case and 5.7 for the bias towards alcohol fun case.
5
Alcohol units of consumption per year in cohort2,280,000
6
Units of alcohol consumption per death386,244
Calcualted by dividing GBD estimate of alcohol deaths in India by WHO estimates of alcohol consumption in India. For every ~135,000 beers drunk, someone dies.
7
8
Deaths in the cohort caused by alcohol consumed in that year
6
9
10
Increase in happiness in the moment from one drinking session
3.65
Reported happiness score out of 100 in Geiger 2016 asspcoated with having recently drunk. That study seems high quality to me. It controls for for fixed effects within people (i.e. it adjusts for whether generally happier people drank more and whether people were happy in the last check in before they started drinking) so shouldn't be confounded by unobserved characteristics that affect both happiness and how much people drink. I expect there are limitations of the study I didn't find on a quick skim.
11
Proportion of a day that increase in happiness lasts18.75%
Guess. Participants in the study were pinged at 2 random times per day to ask what they were just doing and how happy they are out of 100. Participants who said they were recently drinking were 3.65 points happier (after controls; >10 points before controls). Maybe start drinking at 8pm on average when I do drink and go to sleep at midnight. Assuming the happiness of drinking persists until sleep. 4 hours is ~25% of the day.
12
Days in a year365
13
Increase in annual happiness from one drinking session (score out of 100)
0.001875
14
How many units per drinking session?6
Guess. 6 = equivalent to ~3 glasses of wine or 3 pints of lager (https://www.nhs.uk/live-well/alcohol-support/binge-drinking-effects/). Expect this will be a heavy-tailed distribution but haven't checked that.
15
Increase in annualized happiness per unit of alcohol (score out of 100)
0.0003125
e.g. if this was 1, it would mean drinking one unit of alcohol per year increased reported happiness from 70->71 for one year
16
17
Increase in annualized happiness in cohort from drinking
712.5
If this was 10,000, it'd imply that each person in the cohort increases their average happiness over the year from 70->71
18
19
Value of a life saved from alcohol57
Alcohol deaths tends to be concentrated in older age groups.
20
Value of increasing annualized happiness by 1 point0.06
We value a life at 100. So let's say that's 30 DALYs on average. A DALY is equal to a year of life, and let's say (conservative guess) that the difference between 40 and 100 in happiness scores is equal to one DALY. So each DALY is worth 60 happiness points for 1 year. So 1 happiness point is worth 1/60 DALY = 1/1800 of a life = 1/18 of doubling consumption for a year (ie units of value).
21
22
Value from preventing deaths336
23
Disvalue from preventing enjoyment40
24
What % of value from preventing deaths wiped out by enjoyment?
12%
25
26
Verbally: Of 10,000 people in a year, 6 will die because they drink. But drinking increases average moment to moment annual satisfaction of 10k people by 0.1 points / 100. I guess we'd value preventing 6 deaths from alcohol ~4x more than increasing 1,000 people's moment to moment annual satisfaction by 0.1
27
28
29
30
31
32
33
34
35
36
37
38
39
40
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