Info from FEM (note that FEM has especially low cost per person): Freedom Radio reaches 5.6 million listeners (based on third party eval, freedom radio says they reach 10 million), 40% of which are women of reproductive age. I used estimates based off of their Kano campaign only.

This is based on RCT in Burkina Faso, from DMI

I discounted for generalisability (discount is 45%): I am uncertain about this value. Survey data is from women who live in rural areas, <5km from health facility, in villages without electricty access. I think these factors likely limit generalisability. (Note though that the paper found fairly similar results using admin data, which is more representative)

From Kano state data, but using time trend from world bank MMR Nigeria (here)

The Kano data is from 2014 (1025 per 100,000), and it seems likely that MMR has decreased since then. In world bank data, it has decreased by 2.75%, so I decreased Kano MMR by this amount

https://www.sciencedirect.com/science/article/pii/S2214109X14703481 (see fig 2 for Nigeria; 69/ 33317)

https://www.sciencedirect.com/science/article/pii/S2214109X14703481

https://www.sciencedirect.com/science/article/pii/S2214109X14703481

This is the average disability weight from rectovaginal fistula and vesicovaginal fistula

This is quite rough. From paper, most of the women who use contraception at baseline were using implants (~51%), while some were using injectables (~37%) and others were using pills (~9%). I assumed that the average implant lasts around 3 years, and that the women on injectables or pills would use them for around 1 year, then did a weighted average.

See the 'estimated births/ deaths averted' sheet

sanity check: the Glennerster paper says that births reduced by ~7% (the 10% figure is for one age group only). 2240000 (no of women reached in total) x 0.19 (birth rate of woman in Kano) x 0.07 (effect size Glennerster) = 31792 births avoided.

See the 'estimated births/ deaths averted' sheet (then has generalisability discount)

Fistula prevalence x percentage of fistulas that are due to pregnancy/ average number of children for a person with a fistula

Chance of fistula per pregnancy x number of averted pregnancies

From email with Anna Christina. This is their cost from campaign in Kano.

Using 1 maternal life = 41.4 DALY-equivalents

I think is is unlikely (but possible) that FEM has unexpected negative effect. This is a subjective estimate.

167 value points for averting 1 > 5 death, 3.8 value points per DALY

This includes discount for probability that FEM will drop a location, and discount for generalisability from Burkina Faso trial

I put the discount rate higher than usual, because i think contraceptive access/ knowledge will increase even without this intervention

Using 1 maternal life = 41.4 DALY-equivalents

I assume that within 10 years time, even without this intervnetion there would be widespread contraceptive knowledge/ access

Subjective estimate. I think it is unlikely, but not very unlikely, that this campaign would have impacts that last for 4 years and asre strongly diminuishing year by year.

Positive scenario 3: effects over 8 years (if contraceptive accesss/ knowledge became new social norm)

Note: I think it's unlikely, but not impossibly so, that new contraceptive knowledge/ behavior establishes self as new social norm